hog.cpp

/*M///////////////////////////////////////////////////////////////////////////////////////
//
//  IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
//  By downloading, copying, installing or using the software you agree to this license.
//  If you do not agree to this license, do not download, install,
//  copy or use the software.
//
//
//                           License Agreement
//                For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2009, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
//   * Redistribution's of source code must retain the above copyright notice,
//     this list of conditions and the following disclaimer.
//
//   * Redistribution's in binary form must reproduce the above copyright notice,
//     this list of conditions and the following disclaimer in the documentation
//     and/or other materials provided with the distribution.
//
//   * The name of the copyright holders may not be used to endorse or promote products
//     derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/

#include "precomp.hpp"
#include "cascadedetect.hpp"
#include "opencv2/core/core_c.h"
#include "opencl_kernels_objdetect.hpp"

#include <cstdio>
#include <iterator>
#include <limits>

/****************************************************************************************\
      The code below is implementation of HOG (Histogram-of-Oriented Gradients)
      descriptor and object detection, introduced by Navneet Dalal and Bill Triggs.

      The computed feature vectors are compatible with the
      INRIA Object Detection and Localization Toolkit
      (http://pascal.inrialpes.fr/soft/olt/)
\****************************************************************************************/

namespace cv
{

#define NTHREADS 256

static int numPartsWithin(int size, int part_size, int stride)
{
    return (size - part_size + stride) / stride;
}

static Size numPartsWithin(cv::Size size, cv::Size part_size,
                                                cv::Size stride)
{
    return Size(numPartsWithin(size.width, part_size.width, stride.width),
        numPartsWithin(size.height, part_size.height, stride.height));
}

static size_t getBlockHistogramSize(Size block_size, Size cell_size, int nbins)
{
    Size cells_per_block = Size(block_size.width / cell_size.width,
        block_size.height / cell_size.height);
    return (size_t)(nbins * cells_per_block.area());
}

size_t HOGDescriptor::getDescriptorSize() const
{
    CV_Assert(blockSize.width % cellSize.width == 0 &&
        blockSize.height % cellSize.height == 0);
    CV_Assert((winSize.width - blockSize.width) % blockStride.width == 0 &&
        (winSize.height - blockSize.height) % blockStride.height == 0 );

    return (size_t)nbins*
        (blockSize.width/cellSize.width)*
        (blockSize.height/cellSize.height)*
        ((winSize.width - blockSize.width)/blockStride.width + 1)*
        ((winSize.height - blockSize.height)/blockStride.height + 1);
}

double HOGDescriptor::getWinSigma() const
{
    return winSigma > 0 ? winSigma : (blockSize.width + blockSize.height)/8.;
}

bool HOGDescriptor::checkDetectorSize() const
{
    size_t detectorSize = svmDetector.size(), descriptorSize = getDescriptorSize();
    return detectorSize == 0 ||
        detectorSize == descriptorSize ||
        detectorSize == descriptorSize + 1;
}

void HOGDescriptor::setSVMDetector(InputArray _svmDetector)
{
    _svmDetector.getMat().convertTo(svmDetector, CV_32F);
    CV_Assert(checkDetectorSize());

    Mat detector_reordered(1, (int)svmDetector.size(), CV_32FC1);

    size_t block_hist_size = getBlockHistogramSize(blockSize, cellSize, nbins);
    cv::Size blocks_per_img = numPartsWithin(winSize, blockSize, blockStride);

    for (int i = 0; i < blocks_per_img.height; ++i)
        for (int j = 0; j < blocks_per_img.width; ++j)
        {
            const float *src = &svmDetector[0] + (j * blocks_per_img.height + i) * block_hist_size;
            float *dst = detector_reordered.ptr<float>() + (i * blocks_per_img.width + j) * block_hist_size;
            for (size_t k = 0; k < block_hist_size; ++k)
                dst[k] = src[k];
        }
    size_t descriptor_size = getDescriptorSize();
    free_coef = svmDetector.size() > descriptor_size ? svmDetector[descriptor_size] : 0;
    detector_reordered.copyTo(oclSvmDetector);
}

#define CV_TYPE_NAME_HOG_DESCRIPTOR "opencv-object-detector-hog"

bool HOGDescriptor::read(FileNode& obj)
{
    CV_Assert(!obj["winSize"].empty());

    if( !obj.isMap() )
        return false;
    FileNodeIterator it = obj["winSize"].begin();
    it >> winSize.width >> winSize.height;
    it = obj["blockSize"].begin();
    it >> blockSize.width >> blockSize.height;
    it = obj["blockStride"].begin();
    it >> blockStride.width >> blockStride.height;
    it = obj["cellSize"].begin();
    it >> cellSize.width >> cellSize.height;
    obj["nbins"] >> nbins;
    obj["derivAperture"] >> derivAperture;
    obj["winSigma"] >> winSigma;
    obj["histogramNormType"] >> histogramNormType;
    obj["L2HysThreshold"] >> L2HysThreshold;
    obj["gammaCorrection"] >> gammaCorrection;
    obj["nlevels"] >> nlevels;
    if (obj["signedGradient"].empty())
        signedGradient = false;
    else
        obj["signedGradient"] >> signedGradient;

    FileNode vecNode = obj["SVMDetector"];
    if( vecNode.isSeq() )
    {
        std::vector<float> _svmDetector;
        vecNode >> _svmDetector;
        setSVMDetector(_svmDetector);
    }
    return true;
}

void HOGDescriptor::write(FileStorage& fs, const String& objName) const
{
    if( !objName.empty() )
        fs << objName;

    fs << "{" CV_TYPE_NAME_HOG_DESCRIPTOR
       << "winSize" << winSize
       << "blockSize" << blockSize
       << "blockStride" << blockStride
       << "cellSize" << cellSize
       << "nbins" << nbins
       << "derivAperture" << derivAperture
       << "winSigma" << getWinSigma()
       << "histogramNormType" << histogramNormType
       << "L2HysThreshold" << L2HysThreshold
       << "gammaCorrection" << gammaCorrection
       << "nlevels" << nlevels
       << "signedGradient" << signedGradient;
    if( !svmDetector.empty() )
        fs << "SVMDetector" << svmDetector;
    fs << "}";
}

bool HOGDescriptor::load(const String& filename, const String& objname)
{
    FileStorage fs(filename, FileStorage::READ);
    FileNode obj = !objname.empty() ? fs[objname] : fs.getFirstTopLevelNode();
    return read(obj);
}

void HOGDescriptor::save(const String& filename, const String& objName) const
{
    FileStorage fs(filename, FileStorage::WRITE);
    write(fs, !objName.empty() ? objName : FileStorage::getDefaultObjectName(filename));
}

void HOGDescriptor::copyTo(HOGDescriptor& c) const
{
    c.winSize = winSize;
    c.blockSize = blockSize;
    c.blockStride = blockStride;
    c.cellSize = cellSize;
    c.nbins = nbins;
    c.derivAperture = derivAperture;
    c.winSigma = winSigma;
    c.histogramNormType = histogramNormType;
    c.L2HysThreshold = L2HysThreshold;
    c.gammaCorrection = gammaCorrection;
    c.setSVMDetector(svmDetector);
    c.nlevels = nlevels;
    c.signedGradient = signedGradient;
}

#if CV_NEON
// replace of _mm_set_ps
inline float32x4_t vsetq_f32(float f0, float f1, float f2, float f3)
{
    float32x4_t a = vdupq_n_f32(f0);
    a = vsetq_lane_f32(f1, a, 1);
    a = vsetq_lane_f32(f2, a, 2);
    a = vsetq_lane_f32(f3, a, 3);
    return a;
}
#endif
void HOGDescriptor::computeGradient(InputArray _img, InputOutputArray _grad, InputOutputArray _qangle,
    Size paddingTL, Size paddingBR) const
{
    CV_INSTRUMENT_REGION();

    Mat img = _img.getMat();
    CV_Assert(!img.empty());
    CV_Assert( img.type() == CV_8U || img.type() == CV_8UC3 );

    Size gradsize(img.cols + paddingTL.width + paddingBR.width,
        img.rows + paddingTL.height + paddingBR.height);
    _grad.create(gradsize, CV_32FC2);  // <magnitude*(1-alpha), magnitude*alpha>
    _qangle.create(gradsize, CV_8UC2); // [0..nbins-1] - quantized gradient orientation
    Mat grad = _grad.getMat();
    Mat qangle = _qangle.getMat();

    Size wholeSize;
    Point roiofs;
    img.locateROI(wholeSize, roiofs);

    int i, x, y;
    int cn = img.channels();

    Mat_<float> _lut(1, 256);
    const float* const lut = &_lut(0,0);
#if CV_SSE2
    const int indices[] = { 0, 1, 2, 3 };
    __m128i idx = _mm_loadu_si128((const __m128i*)indices);
    __m128i ifour = _mm_set1_epi32(4);

    float* const _data = &_lut(0, 0);
    if( gammaCorrection )
        for( i = 0; i < 256; i += 4 )
        {
            _mm_storeu_ps(_data + i, _mm_sqrt_ps(_mm_cvtepi32_ps(idx)));
            idx = _mm_add_epi32(idx, ifour);
        }
    else
        for( i = 0; i < 256; i += 4 )
        {
            _mm_storeu_ps(_data + i, _mm_cvtepi32_ps(idx));
            idx = _mm_add_epi32(idx, ifour);
        }
#elif CV_NEON
    const int indices[] = { 0, 1, 2, 3 };
    uint32x4_t idx = *(uint32x4_t*)indices;
    uint32x4_t ifour = vdupq_n_u32(4);

    float* const _data = &_lut(0, 0);
    if( gammaCorrection )
        for( i = 0; i < 256; i++ )
            _lut(0,i) = std::sqrt((float)i);
    else
        for( i = 0; i < 256; i += 4 )
        {
            vst1q_f32(_data + i, vcvtq_f32_u32(idx));
            idx = vaddq_u32 (idx, ifour);
        }
#else
    if( gammaCorrection )
        for( i = 0; i < 256; i++ )
            _lut(0,i) = std::sqrt((float)i);
    else
        for( i = 0; i < 256; i++ )
            _lut(0,i) = (float)i;
#endif

    AutoBuffer<int> mapbuf(gradsize.width + gradsize.height + 4);
    int* xmap = mapbuf.data() + 1;
    int* ymap = xmap + gradsize.width + 2;

    const int borderType = (int)BORDER_REFLECT_101;

    for( x = -1; x < gradsize.width + 1; x++ )
        xmap[x] = borderInterpolate(x - paddingTL.width + roiofs.x,
        wholeSize.width, borderType) - roiofs.x;
    for( y = -1; y < gradsize.height + 1; y++ )
        ymap[y] = borderInterpolate(y - paddingTL.height + roiofs.y,
        wholeSize.height, borderType) - roiofs.y;

    // x- & y- derivatives for the whole row
    int width = gradsize.width;
    AutoBuffer<float> _dbuf(width*4);
    float* const dbuf = _dbuf.data();
    Mat Dx(1, width, CV_32F, dbuf);
    Mat Dy(1, width, CV_32F, dbuf + width);
    Mat Mag(1, width, CV_32F, dbuf + width*2);
    Mat Angle(1, width, CV_32F, dbuf + width*3);

    if (cn == 3)
    {
        int end = gradsize.width + 2;
        xmap -= 1, x = 0;
#if CV_SSE2
        for ( ; x <= end - 4; x += 4)
        {
            __m128i mul_res = _mm_loadu_si128((const __m128i*)(xmap + x));
            mul_res = _mm_add_epi32(_mm_add_epi32(mul_res, mul_res), mul_res); // multiply by 3
            _mm_storeu_si128((__m128i*)(xmap + x), mul_res);
        }
#elif CV_NEON
        int32x4_t ithree = vdupq_n_s32(3);
        for ( ; x <= end - 4; x += 4)
            vst1q_s32(xmap + x, vmulq_s32(ithree, vld1q_s32(xmap + x)));
#endif
        for ( ; x < end; ++x)
            xmap[x] *= 3;
        xmap += 1;
    }

    float angleScale = signedGradient ? (float)(nbins/(2.0*CV_PI)) : (float)(nbins/CV_PI);
    for( y = 0; y < gradsize.height; y++ )
    {
        const uchar* imgPtr  = img.ptr(ymap[y]);
        //In case subimage is used ptr() generates an assert for next and prev rows
        //(see http://code.opencv.org/issues/4149)
        const uchar* prevPtr = img.data + img.step*ymap[y-1];
        const uchar* nextPtr = img.data + img.step*ymap[y+1];

        float* gradPtr = grad.ptr<float>(y);
        uchar* qanglePtr = qangle.ptr(y);

        if( cn == 1 )
        {
            for( x = 0; x < width; x++ )
            {
                int x1 = xmap[x];
                dbuf[x] = (float)(lut[imgPtr[xmap[x+1]]] - lut[imgPtr[xmap[x-1]]]);
                dbuf[width + x] = (float)(lut[nextPtr[x1]] - lut[prevPtr[x1]]);
            }
        }
        else
        {
            x = 0;
#if CV_SSE2
            for( ; x <= width - 4; x += 4 )
            {
                int x0 = xmap[x], x1 = xmap[x+1], x2 = xmap[x+2], x3 = xmap[x+3];
                typedef const uchar* const T;
                T p02 = imgPtr + xmap[x+1], p00 = imgPtr + xmap[x-1];
                T p12 = imgPtr + xmap[x+2], p10 = imgPtr + xmap[x];
                T p22 = imgPtr + xmap[x+3], p20 = p02;
                T p32 = imgPtr + xmap[x+4], p30 = p12;

                __m128 _dx0 = _mm_sub_ps(_mm_set_ps(lut[p32[0]], lut[p22[0]], lut[p12[0]], lut[p02[0]]),
                                         _mm_set_ps(lut[p30[0]], lut[p20[0]], lut[p10[0]], lut[p00[0]]));
                __m128 _dx1 = _mm_sub_ps(_mm_set_ps(lut[p32[1]], lut[p22[1]], lut[p12[1]], lut[p02[1]]),
                                         _mm_set_ps(lut[p30[1]], lut[p20[1]], lut[p10[1]], lut[p00[1]]));
                __m128 _dx2 = _mm_sub_ps(_mm_set_ps(lut[p32[2]], lut[p22[2]], lut[p12[2]], lut[p02[2]]),
                                         _mm_set_ps(lut[p30[2]], lut[p20[2]], lut[p10[2]], lut[p00[2]]));

                __m128 _dy0 = _mm_sub_ps(_mm_set_ps(lut[nextPtr[x3]], lut[nextPtr[x2]], lut[nextPtr[x1]], lut[nextPtr[x0]]),
                                         _mm_set_ps(lut[prevPtr[x3]], lut[prevPtr[x2]], lut[prevPtr[x1]], lut[prevPtr[x0]]));
                __m128 _dy1 = _mm_sub_ps(_mm_set_ps(lut[nextPtr[x3+1]], lut[nextPtr[x2+1]], lut[nextPtr[x1+1]], lut[nextPtr[x0+1]]),
                                         _mm_set_ps(lut[prevPtr[x3+1]], lut[prevPtr[x2+1]], lut[prevPtr[x1+1]], lut[prevPtr[x0+1]]));
                __m128 _dy2 = _mm_sub_ps(_mm_set_ps(lut[nextPtr[x3+2]], lut[nextPtr[x2+2]], lut[nextPtr[x1+2]], lut[nextPtr[x0+2]]),
                                         _mm_set_ps(lut[prevPtr[x3+2]], lut[prevPtr[x2+2]], lut[prevPtr[x1+2]], lut[prevPtr[x0+2]]));

                __m128 _mag0 = _mm_add_ps(_mm_mul_ps(_dx0, _dx0), _mm_mul_ps(_dy0, _dy0));
                __m128 _mag1 = _mm_add_ps(_mm_mul_ps(_dx1, _dx1), _mm_mul_ps(_dy1, _dy1));
                __m128 _mag2 = _mm_add_ps(_mm_mul_ps(_dx2, _dx2), _mm_mul_ps(_dy2, _dy2));

                __m128 mask = _mm_cmpgt_ps(_mag2, _mag1);
                _dx2 = _mm_or_ps(_mm_and_ps(_dx2, mask), _mm_andnot_ps(mask, _dx1));
                _dy2 = _mm_or_ps(_mm_and_ps(_dy2, mask), _mm_andnot_ps(mask, _dy1));

                mask = _mm_cmpgt_ps(_mm_max_ps(_mag2, _mag1), _mag0);
                _dx2 = _mm_or_ps(_mm_and_ps(_dx2, mask), _mm_andnot_ps(mask, _dx0));
                _dy2 = _mm_or_ps(_mm_and_ps(_dy2, mask), _mm_andnot_ps(mask, _dy0));

                _mm_storeu_ps(dbuf + x, _dx2);
                _mm_storeu_ps(dbuf + x + width, _dy2);
            }
#elif CV_NEON
            for( ; x <= width - 4; x += 4 )
            {
                int x0 = xmap[x], x1 = xmap[x+1], x2 = xmap[x+2], x3 = xmap[x+3];
                typedef const uchar* const T;
                T p02 = imgPtr + xmap[x+1], p00 = imgPtr + xmap[x-1];
                T p12 = imgPtr + xmap[x+2], p10 = imgPtr + xmap[x];
                T p22 = imgPtr + xmap[x+3], p20 = p02;
                T p32 = imgPtr + xmap[x+4], p30 = p12;

                float32x4_t _dx0 = vsubq_f32(vsetq_f32(lut[p02[0]], lut[p12[0]], lut[p22[0]], lut[p32[0]]),
                                             vsetq_f32(lut[p00[0]], lut[p10[0]], lut[p20[0]], lut[p30[0]]));
                float32x4_t _dx1 = vsubq_f32(vsetq_f32(lut[p02[1]], lut[p12[1]], lut[p22[1]], lut[p32[1]]),
                                             vsetq_f32(lut[p00[1]], lut[p10[1]], lut[p20[1]], lut[p30[1]]));
                float32x4_t _dx2 = vsubq_f32(vsetq_f32(lut[p02[2]], lut[p12[2]], lut[p22[2]], lut[p32[2]]),
                                             vsetq_f32(lut[p00[2]], lut[p10[2]], lut[p20[2]], lut[p30[2]]));

                float32x4_t _dy0 = vsubq_f32(vsetq_f32(lut[nextPtr[x0]], lut[nextPtr[x1]], lut[nextPtr[x2]], lut[nextPtr[x3]]),
                                             vsetq_f32(lut[prevPtr[x0]], lut[prevPtr[x1]], lut[prevPtr[x2]], lut[prevPtr[x3]]));
                float32x4_t _dy1 = vsubq_f32(vsetq_f32(lut[nextPtr[x0+1]], lut[nextPtr[x1+1]], lut[nextPtr[x2+1]], lut[nextPtr[x3+1]]),
                                             vsetq_f32(lut[prevPtr[x0+1]], lut[prevPtr[x1+1]], lut[prevPtr[x2+1]], lut[prevPtr[x3+1]]));
                float32x4_t _dy2 = vsubq_f32(vsetq_f32(lut[nextPtr[x0+2]], lut[nextPtr[x1+2]], lut[nextPtr[x2+2]], lut[nextPtr[x3+2]]),
                                             vsetq_f32(lut[prevPtr[x0+2]], lut[prevPtr[x1+2]], lut[prevPtr[x2+2]], lut[prevPtr[x3+2]]));

                float32x4_t _mag0 = vaddq_f32(vmulq_f32(_dx0, _dx0), vmulq_f32(_dy0, _dy0));
                float32x4_t _mag1 = vaddq_f32(vmulq_f32(_dx1, _dx1), vmulq_f32(_dy1, _dy1));
                float32x4_t _mag2 = vaddq_f32(vmulq_f32(_dx2, _dx2), vmulq_f32(_dy2, _dy2));

                uint32x4_t mask = vcgtq_f32(_mag2, _mag1);
                _dx2 = vbslq_f32(mask, _dx2, _dx1);
                _dy2 = vbslq_f32(mask, _dy2, _dy1);

                mask = vcgtq_f32(vmaxq_f32(_mag2, _mag1), _mag0);
                _dx2 = vbslq_f32(mask, _dx2, _dx0);
                _dy2 = vbslq_f32(mask, _dy2, _dy0);

                vst1q_f32(dbuf + x, _dx2);
                vst1q_f32(dbuf + x + width, _dy2);
            }
#endif
            for( ; x < width; x++ )
            {
                int x1 = xmap[x];
                float dx0, dy0, dx, dy, mag0, mag;
                const uchar* p2 = imgPtr + xmap[x+1];
                const uchar* p0 = imgPtr + xmap[x-1];

                dx0 = lut[p2[2]] - lut[p0[2]];
                dy0 = lut[nextPtr[x1+2]] - lut[prevPtr[x1+2]];
                mag0 = dx0*dx0 + dy0*dy0;

                dx = lut[p2[1]] - lut[p0[1]];
                dy = lut[nextPtr[x1+1]] - lut[prevPtr[x1+1]];
                mag = dx*dx + dy*dy;
                if( mag0 < mag )
                {
                    dx0 = dx;
                    dy0 = dy;
                    mag0 = mag;
                }

                dx = lut[p2[0]] - lut[p0[0]];
                dy = lut[nextPtr[x1]] - lut[prevPtr[x1]];
                mag = dx*dx + dy*dy;
                if( mag0 < mag )
                {
                    dx0 = dx;
                    dy0 = dy;
                    mag0 = mag;
                }

                dbuf[x] = dx0;
                dbuf[x+width] = dy0;
            }
        }

        // computing angles and magnidutes
        cartToPolar( Dx, Dy, Mag, Angle, false );

        // filling the result matrix
        x = 0;
#if CV_SSE2
        __m128 fhalf = _mm_set1_ps(0.5f), fzero = _mm_setzero_ps();
        __m128 _angleScale = _mm_set1_ps(angleScale), fone = _mm_set1_ps(1.0f);
        __m128i ione = _mm_set1_epi32(1), _nbins = _mm_set1_epi32(nbins), izero = _mm_setzero_si128();

        for ( ; x <= width - 4; x += 4)
        {
            int x2 = x << 1;
            __m128 _mag = _mm_loadu_ps(dbuf + x + (width << 1));
            __m128 _angle = _mm_loadu_ps(dbuf + x + width * 3);
            _angle = _mm_sub_ps(_mm_mul_ps(_angleScale, _angle), fhalf);

            __m128 sign = _mm_and_ps(fone, _mm_cmplt_ps(_angle, fzero));
            __m128i _hidx = _mm_cvttps_epi32(_angle);
            _hidx = _mm_sub_epi32(_hidx, _mm_cvtps_epi32(sign));
            _angle = _mm_sub_ps(_angle, _mm_cvtepi32_ps(_hidx));

            __m128 ft0 = _mm_mul_ps(_mag, _mm_sub_ps(fone, _angle));
            __m128 ft1 = _mm_mul_ps(_mag, _angle);
            __m128 ft2 = _mm_unpacklo_ps(ft0, ft1);
            __m128 ft3 = _mm_unpackhi_ps(ft0, ft1);

            _mm_storeu_ps(gradPtr + x2, ft2);
            _mm_storeu_ps(gradPtr + x2 + 4, ft3);

            __m128i mask0 = _mm_sub_epi32(izero, _mm_srli_epi32(_hidx, 31));
            __m128i it0 = _mm_and_si128(mask0, _nbins);
            mask0 = _mm_cmplt_epi32(_hidx, _nbins);
            __m128i it1 = _mm_andnot_si128(mask0, _nbins);
            _hidx = _mm_add_epi32(_hidx, _mm_sub_epi32(it0, it1));

            it0 = _mm_packus_epi16(_mm_packs_epi32(_hidx, izero), izero);
            _hidx = _mm_add_epi32(ione, _hidx);
            _hidx = _mm_and_si128(_hidx, _mm_cmplt_epi32(_hidx, _nbins));
            it1 = _mm_packus_epi16(_mm_packs_epi32(_hidx, izero), izero);
            it0 = _mm_unpacklo_epi8(it0, it1);

            _mm_storel_epi64((__m128i*)(qanglePtr + x2), it0);
        }
#endif
        for( ; x < width; x++ )
        {
            float mag = dbuf[x+width*2], angle = dbuf[x+width*3]*angleScale - 0.5f;
            int hidx = cvFloor(angle);
            angle -= hidx;
            gradPtr[x*2] = mag*(1.f - angle);
            gradPtr[x*2+1] = mag*angle;

            if( hidx < 0 )
                hidx += nbins;
            else if( hidx >= nbins )
                hidx -= nbins;

            CV_Assert( (unsigned)hidx < (unsigned)nbins );

            qanglePtr[x*2] = (uchar)hidx;
            hidx++;
            hidx &= hidx < nbins ? -1 : 0;
            qanglePtr[x*2+1] = (uchar)hidx;
        }
    }
}

struct HOGCache
{
    struct BlockData
    {
        BlockData() :
            histOfs(0), imgOffset()
        { }

        int histOfs;
        Point imgOffset;
    };

    struct PixData
    {
        size_t gradOfs, qangleOfs;
        int histOfs[4];
        float histWeights[4];
        float gradWeight;
    };

    HOGCache();
    HOGCache(const HOGDescriptor* descriptor,
        const Mat& img, const Size& paddingTL, const Size& paddingBR,
        bool useCache, const Size& cacheStride);
    virtual ~HOGCache() { }
    virtual void init(const HOGDescriptor* descriptor,
        const Mat& img, const Size& paddingTL, const Size& paddingBR,
        bool useCache, const Size& cacheStride);

    Size windowsInImage(const Size& imageSize, const Size& winStride) const;
    Rect getWindow(const Size& imageSize, const Size& winStride, int idx) const;

    const float* getBlock(Point pt, float* buf);
    virtual void normalizeBlockHistogram(float* histogram) const;

    std::vector<PixData> pixData;
    std::vector<BlockData> blockData;

    bool useCache;
    std::vector<int> ymaxCached;
    Size winSize;
    Size cacheStride;
    Size nblocks, ncells;
    int blockHistogramSize;
    int count1, count2, count4;
    Point imgoffset;
    Mat_<float> blockCache;
    Mat_<uchar> blockCacheFlags;

    Mat grad, qangle;
    const HOGDescriptor* descriptor;
};

HOGCache::HOGCache() :
    blockHistogramSize(), count1(), count2(), count4()
{
    useCache = false;
    descriptor = 0;
}

HOGCache::HOGCache(const HOGDescriptor* _descriptor,
    const Mat& _img, const Size& _paddingTL, const Size& _paddingBR,
    bool _useCache, const Size& _cacheStride)
{
    init(_descriptor, _img, _paddingTL, _paddingBR, _useCache, _cacheStride);
}

void HOGCache::init(const HOGDescriptor* _descriptor,
    const Mat& _img, const Size& _paddingTL, const Size& _paddingBR,
    bool _useCache, const Size& _cacheStride)
{
    descriptor = _descriptor;
    cacheStride = _cacheStride;
    useCache = _useCache;

    descriptor->computeGradient(_img, grad, qangle, _paddingTL, _paddingBR);
    imgoffset = _paddingTL;

    winSize = descriptor->winSize;
    Size blockSize = descriptor->blockSize;
    Size blockStride = descriptor->blockStride;
    Size cellSize = descriptor->cellSize;
    int i, j, nbins = descriptor->nbins;
    int rawBlockSize = blockSize.width*blockSize.height;

    nblocks = Size((winSize.width - blockSize.width)/blockStride.width + 1,
        (winSize.height - blockSize.height)/blockStride.height + 1);
    ncells = Size(blockSize.width/cellSize.width, blockSize.height/cellSize.height);
    blockHistogramSize = ncells.width*ncells.height*nbins;

    if( useCache )
    {
        Size cacheSize((grad.cols - blockSize.width)/cacheStride.width+1,
            (winSize.height/cacheStride.height)+1);

        blockCache.create(cacheSize.height, cacheSize.width*blockHistogramSize);
        blockCacheFlags.create(cacheSize);

        size_t cacheRows = blockCache.rows;
        ymaxCached.resize(cacheRows);
        for(size_t ii = 0; ii < cacheRows; ii++ )
            ymaxCached[ii] = -1;
    }

    Mat_<float> weights(blockSize);
    float sigma = (float)descriptor->getWinSigma();
    float scale = 1.f/(sigma*sigma*2);

    {
        AutoBuffer<float> di(blockSize.height), dj(blockSize.width);
        float* _di = di.data(), *_dj = dj.data();
        float bh = blockSize.height * 0.5f, bw = blockSize.width * 0.5f;

        i = 0;
    #if CV_SSE2
        const int a[] = { 0, 1, 2, 3 };
        __m128i idx = _mm_loadu_si128((__m128i*)a);
        __m128 _bw = _mm_set1_ps(bw), _bh = _mm_set1_ps(bh);
        __m128i ifour = _mm_set1_epi32(4);

        for (; i <= blockSize.height - 4; i += 4)
        {
            __m128 t = _mm_sub_ps(_mm_cvtepi32_ps(idx), _bh);
            t = _mm_mul_ps(t, t);
            idx = _mm_add_epi32(idx, ifour);
            _mm_storeu_ps(_di + i, t);
        }
    #elif CV_NEON
        const int a[] = { 0, 1, 2, 3 };
        int32x4_t idx = vld1q_s32(a);
        float32x4_t _bw = vdupq_n_f32(bw), _bh = vdupq_n_f32(bh);
        int32x4_t ifour = vdupq_n_s32(4);

        for (; i <= blockSize.height - 4; i += 4)
        {
            float32x4_t t = vsubq_f32(vcvtq_f32_s32(idx), _bh);
            t = vmulq_f32(t, t);
            idx = vaddq_s32(idx, ifour);
            vst1q_f32(_di + i, t);
        }
    #endif
        for ( ; i < blockSize.height; ++i)
        {
            _di[i] = i - bh;
            _di[i] *= _di[i];
        }

        j = 0;
    #if CV_SSE2
        idx = _mm_loadu_si128((__m128i*)a);
        for (; j <= blockSize.width - 4; j += 4)
        {
            __m128 t = _mm_sub_ps(_mm_cvtepi32_ps(idx), _bw);
            t = _mm_mul_ps(t, t);
            idx = _mm_add_epi32(idx, ifour);
            _mm_storeu_ps(_dj + j, t);
        }
    #elif CV_NEON
        idx = vld1q_s32(a);
        for (; j <= blockSize.width - 4; j += 4)
        {
            float32x4_t t = vsubq_f32(vcvtq_f32_s32(idx), _bw);
            t = vmulq_f32(t, t);
            idx = vaddq_s32(idx, ifour);
            vst1q_f32(_dj + j, t);
        }
    #endif
        for ( ; j < blockSize.width; ++j)
        {
            _dj[j] = j - bw;
            _dj[j] *= _dj[j];
        }

        for(i = 0; i < blockSize.height; i++)
            for(j = 0; j < blockSize.width; j++)
                weights(i,j) = std::exp(-(_di[i] + _dj[j])*scale);
    }

    blockData.resize(nblocks.width*nblocks.height);
    pixData.resize(rawBlockSize*3);

    // Initialize 2 lookup tables, pixData & blockData.
    // Here is why:
    //
    // The detection algorithm runs in 4 nested loops (at each pyramid layer):
    //  loop over the windows within the input image
    //    loop over the blocks within each window
    //      loop over the cells within each block
    //        loop over the pixels in each cell
    //
    // As each of the loops runs over a 2-dimensional array,
    // we could get 8(!) nested loops in total, which is very-very slow.
    //
    // To speed the things up, we do the following:
    //   1. loop over windows is unrolled in the HOGDescriptor::{compute|detect} methods;
    //         inside we compute the current search window using getWindow() method.
    //         Yes, it involves some overhead (function call + couple of divisions),
    //         but it's tiny in fact.
    //   2. loop over the blocks is also unrolled. Inside we use pre-computed blockData[j]
    //         to set up gradient and histogram pointers.
    //   3. loops over cells and pixels in each cell are merged
    //       (since there is no overlap between cells, each pixel in the block is processed once)
    //      and also unrolled. Inside we use PixData[k] to access the gradient values and
    //      update the histogram
    //

    count1 = count2 = count4 = 0;
    for( j = 0; j < blockSize.width; j++ )
        for( i = 0; i < blockSize.height; i++ )
        {
            PixData* data = 0;
            float cellX = (j+0.5f)/cellSize.width - 0.5f;
            float cellY = (i+0.5f)/cellSize.height - 0.5f;
            int icellX0 = cvFloor(cellX);
            int icellY0 = cvFloor(cellY);
            int icellX1 = icellX0 + 1, icellY1 = icellY0 + 1;
            cellX -= icellX0;
            cellY -= icellY0;

            if( (unsigned)icellX0 < (unsigned)ncells.width &&
               (unsigned)icellX1 < (unsigned)ncells.width )
            {
                if( (unsigned)icellY0 < (unsigned)ncells.height &&
                   (unsigned)icellY1 < (unsigned)ncells.height )
                {
                    data = &pixData[rawBlockSize*2 + (count4++)];
                    data->histOfs[0] = (icellX0*ncells.height + icellY0)*nbins;
                    data->histWeights[0] = (1.f - cellX)*(1.f - cellY);
                    data->histOfs[1] = (icellX1*ncells.height + icellY0)*nbins;
                    data->histWeights[1] = cellX*(1.f - cellY);
                    data->histOfs[2] = (icellX0*ncells.height + icellY1)*nbins;
                    data->histWeights[2] = (1.f - cellX)*cellY;
                    data->histOfs[3] = (icellX1*ncells.height + icellY1)*nbins;
                    data->histWeights[3] = cellX*cellY;
                }
                else
                {
                    data = &pixData[rawBlockSize + (count2++)];
                    if( (unsigned)icellY0 < (unsigned)ncells.height )
                    {
                        icellY1 = icellY0;
                        cellY = 1.f - cellY;
                    }
                    data->histOfs[0] = (icellX0*ncells.height + icellY1)*nbins;
                    data->histWeights[0] = (1.f - cellX)*cellY;
                    data->histOfs[1] = (icellX1*ncells.height + icellY1)*nbins;
                    data->histWeights[1] = cellX*cellY;
                    data->histOfs[2] = data->histOfs[3] = 0;
                    data->histWeights[2] = data->histWeights[3] = 0;
                }
            }
            else
            {
                if( (unsigned)icellX0 < (unsigned)ncells.width )
                {
                    icellX1 = icellX0;
                    cellX = 1.f - cellX;
                }

                if( (unsigned)icellY0 < (unsigned)ncells.height &&
                   (unsigned)icellY1 < (unsigned)ncells.height )
                {
                    data = &pixData[rawBlockSize + (count2++)];
                    data->histOfs[0] = (icellX1*ncells.height + icellY0)*nbins;
                    data->histWeights[0] = cellX*(1.f - cellY);
                    data->histOfs[1] = (icellX1*ncells.height + icellY1)*nbins;
                    data->histWeights[1] = cellX*cellY;
                    data->histOfs[2] = data->histOfs[3] = 0;
                    data->histWeights[2] = data->histWeights[3] = 0;
                }
                else
                {
                    data = &pixData[count1++];
                    if( (unsigned)icellY0 < (unsigned)ncells.height )
                    {
                        icellY1 = icellY0;
                        cellY = 1.f - cellY;
                    }
                    data->histOfs[0] = (icellX1*ncells.height + icellY1)*nbins;
                    data->histWeights[0] = cellX*cellY;
                    data->histOfs[1] = data->histOfs[2] = data->histOfs[3] = 0;
                    data->histWeights[1] = data->histWeights[2] = data->histWeights[3] = 0;
                }
            }
            data->gradOfs = (grad.cols*i + j)*2;
            data->qangleOfs = (qangle.cols*i + j)*2;
            data->gradWeight = weights(i,j);
        }

    assert( count1 + count2 + count4 == rawBlockSize );
    // defragment pixData
    for( j = 0; j < count2; j++ )
        pixData[j + count1] = pixData[j + rawBlockSize];
    for( j = 0; j < count4; j++ )
        pixData[j + count1 + count2] = pixData[j + rawBlockSize*2];
    count2 += count1;
    count4 += count2;

    // initialize blockData
    for( j = 0; j < nblocks.width; j++ )
        for( i = 0; i < nblocks.height; i++ )
        {
            BlockData& data = blockData[j*nblocks.height + i];
            data.histOfs = (j*nblocks.height + i)*blockHistogramSize;
            data.imgOffset = Point(j*blockStride.width,i*blockStride.height);
        }
}

const float* HOGCache::getBlock(Point pt, float* buf)
{
    float* blockHist = buf;
    assert(descriptor != 0);

//    Size blockSize = descriptor->blockSize;
    pt += imgoffset;

//    CV_Assert( (unsigned)pt.x <= (unsigned)(grad.cols - blockSize.width) &&
//        (unsigned)pt.y <= (unsigned)(grad.rows - blockSize.height) );

    if( useCache )
    {
        CV_Assert( pt.x % cacheStride.width == 0 &&
                   pt.y % cacheStride.height == 0 );
        Point cacheIdx(pt.x/cacheStride.width,
                       (pt.y/cacheStride.height) % blockCache.rows);
        if( pt.y != ymaxCached[cacheIdx.y] )
        {
            Mat_<uchar> cacheRow = blockCacheFlags.row(cacheIdx.y);
            cacheRow = (uchar)0;
            ymaxCached[cacheIdx.y] = pt.y;
        }

        blockHist = &blockCache[cacheIdx.y][cacheIdx.x*blockHistogramSize];
        uchar& computedFlag = blockCacheFlags(cacheIdx.y, cacheIdx.x);
        if( computedFlag != 0 )
            return blockHist;
        computedFlag = (uchar)1; // set it at once, before actual computing
    }

    int k, C1 = count1, C2 = count2, C4 = count4;
    const float* gradPtr = grad.ptr<float>(pt.y) + pt.x*2;
    const uchar* qanglePtr = qangle.ptr(pt.y) + pt.x*2;

//    CV_Assert( blockHist != 0 );
    memset(blockHist, 0, sizeof(float) * blockHistogramSize);

    const PixData* _pixData = &pixData[0];

    for( k = 0; k < C1; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        float w = pk.gradWeight*pk.histWeights[0];
        const uchar* h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        float* hist = blockHist + pk.histOfs[0];
        float t0 = hist[h0] + a[0]*w;
        float t1 = hist[h1] + a[1]*w;
        hist[h0] = t0; hist[h1] = t1;
    }

#if CV_SSE2
    float hist0[4], hist1[4];
    for( ; k < C2; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        const uchar* const h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        __m128 _a0 = _mm_set1_ps(a[0]), _a1 = _mm_set1_ps(a[1]);
        __m128 _w = _mm_mul_ps(_mm_set1_ps(pk.gradWeight), _mm_loadu_ps(pk.histWeights));
        __m128 _t0 = _mm_mul_ps(_a0, _w), _t1 = _mm_mul_ps(_a1, _w);

        _mm_storeu_ps(hist0, _t0);
        _mm_storeu_ps(hist1, _t1);

        float* hist = blockHist + pk.histOfs[0];
        float t0 = hist[h0] + hist0[0];
        float t1 = hist[h1] + hist1[0];
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[1];
        t0 = hist[h0] + hist0[1];
        t1 = hist[h1] + hist1[1];
        hist[h0] = t0; hist[h1] = t1;
    }
#elif CV_NEON
    float hist0[4], hist1[4];
    for( ; k < C2; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        const uchar* const h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        float32x4_t _a0 = vdupq_n_f32(a[0]), _a1 = vdupq_n_f32(a[1]);
        float32x4_t _w = vmulq_f32(vdupq_n_f32(pk.gradWeight), vld1q_f32(pk.histWeights));

        float32x4_t _h0 = vsetq_f32((blockHist + pk.histOfs[0])[h0], (blockHist + pk.histOfs[1])[h0], 0,  0);
        float32x4_t _h1 = vsetq_f32((blockHist + pk.histOfs[0])[h1], (blockHist + pk.histOfs[1])[h1], 0,  0);

        float32x4_t _t0 = vmlaq_f32(_h0, _a0, _w), _t1 = vmlaq_f32(_h1, _a1, _w);
        vst1q_f32(hist0, _t0);
        vst1q_f32(hist1, _t1);

        (blockHist + pk.histOfs[0])[h0] = hist0[0];
        (blockHist + pk.histOfs[1])[h0] = hist0[1];

        (blockHist + pk.histOfs[0])[h1] = hist1[0];
        (blockHist + pk.histOfs[1])[h1] = hist1[1];
    }
#else
    for( ; k < C2; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        float w, t0, t1, a0 = a[0], a1 = a[1];
        const uchar* const h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        float* hist = blockHist + pk.histOfs[0];
        w = pk.gradWeight*pk.histWeights[0];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[1];
        w = pk.gradWeight*pk.histWeights[1];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;
    }
#endif

#if CV_SSE2
    for( ; k < C4; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        const uchar* const h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        __m128 _a0 = _mm_set1_ps(a[0]), _a1 = _mm_set1_ps(a[1]);
        __m128 _w = _mm_mul_ps(_mm_set1_ps(pk.gradWeight), _mm_loadu_ps(pk.histWeights));
        __m128 _t0 = _mm_mul_ps(_a0, _w), _t1 = _mm_mul_ps(_a1, _w);

        _mm_storeu_ps(hist0, _t0);
        _mm_storeu_ps(hist1, _t1);

        float* hist = blockHist + pk.histOfs[0];
        float t0 = hist[h0] + hist0[0];
        float t1 = hist[h1] + hist1[0];
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[1];
        t0 = hist[h0] + hist0[1];
        t1 = hist[h1] + hist1[1];
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[2];
        t0 = hist[h0] + hist0[2];
        t1 = hist[h1] + hist1[2];
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[3];
        t0 = hist[h0] + hist0[3];
        t1 = hist[h1] + hist1[3];
        hist[h0] = t0; hist[h1] = t1;

//        __m128 _hist0 = _mm_set_ps((blockHist + pk.histOfs[3])[h0], (blockHist + pk.histOfs[2])[h0],
//            (blockHist + pk.histOfs[1])[h0], (blockHist + pk.histOfs[0])[h0]);
//        __m128 _hist1 = _mm_set_ps((blockHist + pk.histOfs[3])[h1], (blockHist + pk.histOfs[2])[h1],
//            (blockHist + pk.histOfs[1])[h1], (blockHist + pk.histOfs[0])[h1]);
//
//        _hist0 = _mm_add_ps(_t0, _hist0);
//        _hist1 = _mm_add_ps(_t1, _hist1);
//
//        _mm_storeu_ps(hist0, _hist0);
//        _mm_storeu_ps(hist1, _hist1);
//
//        (pk.histOfs[0] + blockHist)[h0] = hist0[0];
//        (pk.histOfs[1] + blockHist)[h0] = hist0[1];
//        (pk.histOfs[2] + blockHist)[h0] = hist0[2];
//        (pk.histOfs[3] + blockHist)[h0] = hist0[3];
//
//        (pk.histOfs[0] + blockHist)[h1] = hist1[0];
//        (pk.histOfs[1] + blockHist)[h1] = hist1[1];
//        (pk.histOfs[2] + blockHist)[h1] = hist1[2];
//        (pk.histOfs[3] + blockHist)[h1] = hist1[3];
    }
#elif CV_NEON
    for( ; k < C4; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* const a = gradPtr + pk.gradOfs;
        const uchar* const h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        float32x4_t _a0 = vdupq_n_f32(a[0]), _a1 = vdupq_n_f32(a[1]);
        float32x4_t _w = vmulq_f32(vdupq_n_f32(pk.gradWeight), vld1q_f32(pk.histWeights));

        float32x4_t _h0 = vsetq_f32((blockHist + pk.histOfs[0])[h0],
                                    (blockHist + pk.histOfs[1])[h0],
                                    (blockHist + pk.histOfs[2])[h0],
                                    (blockHist + pk.histOfs[3])[h0]);
        float32x4_t _h1 = vsetq_f32((blockHist + pk.histOfs[0])[h1],
                                    (blockHist + pk.histOfs[1])[h1],
                                    (blockHist + pk.histOfs[2])[h1],
                                    (blockHist + pk.histOfs[3])[h1]);


        float32x4_t _t0 = vmlaq_f32(_h0, _a0, _w), _t1 = vmlaq_f32(_h1, _a1, _w);
        vst1q_f32(hist0, _t0);
        vst1q_f32(hist1, _t1);

        (blockHist + pk.histOfs[0])[h0] = hist0[0];
        (blockHist + pk.histOfs[1])[h0] = hist0[1];
        (blockHist + pk.histOfs[2])[h0] = hist0[2];
        (blockHist + pk.histOfs[3])[h0] = hist0[3];

        (blockHist + pk.histOfs[0])[h1] = hist1[0];
        (blockHist + pk.histOfs[1])[h1] = hist1[1];
        (blockHist + pk.histOfs[2])[h1] = hist1[2];
        (blockHist + pk.histOfs[3])[h1] = hist1[3];
    }
#else
    for( ; k < C4; k++ )
    {
        const PixData& pk = _pixData[k];
        const float* a = gradPtr + pk.gradOfs;
        float w, t0, t1, a0 = a[0], a1 = a[1];
        const uchar* h = qanglePtr + pk.qangleOfs;
        int h0 = h[0], h1 = h[1];

        float* hist = blockHist + pk.histOfs[0];
        w = pk.gradWeight*pk.histWeights[0];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[1];
        w = pk.gradWeight*pk.histWeights[1];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[2];
        w = pk.gradWeight*pk.histWeights[2];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;

        hist = blockHist + pk.histOfs[3];
        w = pk.gradWeight*pk.histWeights[3];
        t0 = hist[h0] + a0*w;
        t1 = hist[h1] + a1*w;
        hist[h0] = t0; hist[h1] = t1;
    }
#endif

    normalizeBlockHistogram(blockHist);

    return blockHist;
}

void HOGCache::normalizeBlockHistogram(float* _hist) const
{
    float* hist = &_hist[0], sum = 0.0f, partSum[4];
    size_t i = 0, sz = blockHistogramSize;

#if CV_SSE2
    __m128 p0 = _mm_loadu_ps(hist);
    __m128 s = _mm_mul_ps(p0, p0);

    for (i = 4; i <= sz - 4; i += 4)
    {
        p0 = _mm_loadu_ps(hist + i);
        s = _mm_add_ps(s, _mm_mul_ps(p0, p0));
    }
    _mm_storeu_ps(partSum, s);
#elif CV_NEON
    float32x4_t p0 = vld1q_f32(hist);
    float32x4_t s = vmulq_f32(p0, p0);

    for (i = 4; i <= sz - 4; i += 4)
    {
        p0 = vld1q_f32(hist + i);
        s = vaddq_f32(s, vmulq_f32(p0, p0));
    }
    vst1q_f32(partSum, s);
#else
    partSum[0] = 0.0f;
    partSum[1] = 0.0f;
    partSum[2] = 0.0f;
    partSum[3] = 0.0f;
    for ( ; i <= sz - 4; i += 4)
    {
        partSum[0] += hist[i] * hist[i];
        partSum[1] += hist[i+1] * hist[i+1];
        partSum[2] += hist[i+2] * hist[i+2];
        partSum[3] += hist[i+3] * hist[i+3];
    }
#endif
    float t0 = partSum[0] + partSum[1];
    float t1 = partSum[2] + partSum[3];
    sum = t0 + t1;
    for ( ; i < sz; ++i)
        sum += hist[i]*hist[i];

    float scale = 1.f/(std::sqrt(sum)+sz*0.1f), thresh = (float)descriptor->L2HysThreshold;
    i = 0, sum = 0.0f;

#if CV_SSE2
    __m128 _scale = _mm_set1_ps(scale);
    static __m128 _threshold = _mm_set1_ps(thresh);

    __m128 p = _mm_mul_ps(_scale, _mm_loadu_ps(hist));
    p = _mm_min_ps(p, _threshold);
    s = _mm_mul_ps(p, p);
    _mm_storeu_ps(hist, p);

    for(i = 4 ; i <= sz - 4; i += 4)
    {
        p = _mm_loadu_ps(hist + i);
        p = _mm_mul_ps(p, _scale);
        p = _mm_min_ps(p, _threshold);
        s = _mm_add_ps(s, _mm_mul_ps(p, p));
        _mm_storeu_ps(hist + i, p);
    }

    _mm_storeu_ps(partSum, s);
#elif CV_NEON
    float32x4_t _scale = vdupq_n_f32(scale);
    static float32x4_t _threshold = vdupq_n_f32(thresh);

    float32x4_t p = vmulq_f32(_scale, vld1q_f32(hist));
    p = vminq_f32(p, _threshold);
    s = vmulq_f32(p, p);
    vst1q_f32(hist, p);

    for(i = 4 ; i <= sz - 4; i += 4)
    {
        p = vld1q_f32(hist + i);
        p = vmulq_f32(p, _scale);
        p = vminq_f32(p, _threshold);
        s = vaddq_f32(s, vmulq_f32(p, p));
        vst1q_f32(hist + i, p);
    }

    vst1q_f32(partSum, s);
#else
    partSum[0] = 0.0f;
    partSum[1] = 0.0f;
    partSum[2] = 0.0f;
    partSum[3] = 0.0f;
    for( ; i <= sz - 4; i += 4)
    {
        hist[i] = std::min(hist[i]*scale, thresh);
        hist[i+1] = std::min(hist[i+1]*scale, thresh);
        hist[i+2] = std::min(hist[i+2]*scale, thresh);
        hist[i+3] = std::min(hist[i+3]*scale, thresh);
        partSum[0] += hist[i]*hist[i];
        partSum[1] += hist[i+1]*hist[i+1];
        partSum[2] += hist[i+2]*hist[i+2];
        partSum[3] += hist[i+3]*hist[i+3];
    }
#endif
    t0 = partSum[0] + partSum[1];
    t1 = partSum[2] + partSum[3];
    sum = t0 + t1;
    for( ; i < sz; ++i)
    {
        hist[i] = std::min(hist[i]*scale, thresh);
        sum += hist[i]*hist[i];
    }

    scale = 1.f/(std::sqrt(sum)+1e-3f), i = 0;
#if CV_SSE2
    __m128 _scale2 = _mm_set1_ps(scale);
    for ( ; i <= sz - 4; i += 4)
    {
        __m128 t = _mm_mul_ps(_scale2, _mm_loadu_ps(hist + i));
        _mm_storeu_ps(hist + i, t);
    }
#elif CV_NEON
    float32x4_t _scale2 = vdupq_n_f32(scale);
    for ( ; i <= sz - 4; i += 4)
    {
        float32x4_t t = vmulq_f32(_scale2, vld1q_f32(hist + i));
        vst1q_f32(hist + i, t);
    }
#endif
    for ( ; i < sz; ++i)
        hist[i] *= scale;
}

Size HOGCache::windowsInImage(const Size& imageSize, const Size& winStride) const
{
    return Size((imageSize.width - winSize.width)/winStride.width + 1,
        (imageSize.height - winSize.height)/winStride.height + 1);
}

Rect HOGCache::getWindow(const Size& imageSize, const Size& winStride, int idx) const
{
    int nwindowsX = (imageSize.width - winSize.width)/winStride.width + 1;
    int y = idx / nwindowsX;
    int x = idx - nwindowsX*y;
    return Rect( x*winStride.width, y*winStride.height, winSize.width, winSize.height );
}

static inline int gcd(int a, int b)
{
    if( a < b )
        std::swap(a, b);
    while( b > 0 )
    {
        int r = a % b;
        a = b;
        b = r;
    }
    return a;
}

#ifdef HAVE_OPENCL

static bool ocl_compute_gradients_8UC1(int height, int width, InputArray _img, float angle_scale,
                                       UMat grad, UMat qangle, bool correct_gamma, int nbins)
{
    ocl::Kernel k("compute_gradients_8UC1_kernel", ocl::objdetect::objdetect_hog_oclsrc);
    if(k.empty())
        return false;

    UMat img = _img.getUMat();

    size_t localThreads[3] = { NTHREADS, 1, 1 };
    size_t globalThreads[3] = { (size_t)width, (size_t)height, 1 };
    char correctGamma = (correct_gamma) ? 1 : 0;
    int grad_quadstep = (int)grad.step >> 3;
    int qangle_elem_size = CV_ELEM_SIZE1(qangle.type());
    int qangle_step = (int)qangle.step / (2 * qangle_elem_size);

    int idx = 0;
    idx = k.set(idx, height);
    idx = k.set(idx, width);
    idx = k.set(idx, (int)img.step1());
    idx = k.set(idx, grad_quadstep);
    idx = k.set(idx, qangle_step);
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(img));
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(grad));
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(qangle));
    idx = k.set(idx, angle_scale);
    idx = k.set(idx, correctGamma);
    idx = k.set(idx, nbins);

    return k.run(2, globalThreads, localThreads, false);
}

static bool ocl_computeGradient(InputArray img, UMat grad, UMat qangle, int nbins, Size effect_size, bool gamma_correction, bool signedGradient)
{
    float angleScale = signedGradient ? (float)(nbins/(2.0*CV_PI)) : (float)(nbins/CV_PI);

    return ocl_compute_gradients_8UC1(effect_size.height, effect_size.width, img,
         angleScale, grad, qangle, gamma_correction, nbins);
}

#define CELL_WIDTH 8
#define CELL_HEIGHT 8
#define CELLS_PER_BLOCK_X 2
#define CELLS_PER_BLOCK_Y 2

static bool ocl_compute_hists(int nbins, int block_stride_x, int block_stride_y, int height, int width,
                              UMat grad, UMat qangle, UMat gauss_w_lut, UMat block_hists, size_t block_hist_size)
{
    ocl::Kernel k("compute_hists_lut_kernel", ocl::objdetect::objdetect_hog_oclsrc);
    if(k.empty())
        return false;
    bool is_cpu = cv::ocl::Device::getDefault().type() == cv::ocl::Device::TYPE_CPU;
    cv::String opts;
    if(is_cpu)
       opts = "-D CPU ";
    else
        opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
    k.create("compute_hists_lut_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
    if(k.empty())
        return false;

    int img_block_width = (width - CELLS_PER_BLOCK_X * CELL_WIDTH + block_stride_x)/block_stride_x;
    int img_block_height = (height - CELLS_PER_BLOCK_Y * CELL_HEIGHT + block_stride_y)/block_stride_y;
    int blocks_total = img_block_width * img_block_height;

    int qangle_elem_size = CV_ELEM_SIZE1(qangle.type());
    int grad_quadstep = (int)grad.step >> 2;
    int qangle_step = (int)qangle.step / qangle_elem_size;

    int blocks_in_group = 4;
    size_t localThreads[3] = { (size_t)blocks_in_group * 24, 2, 1 };
    size_t globalThreads[3] = {((img_block_width * img_block_height + blocks_in_group - 1)/blocks_in_group) * localThreads[0], 2, 1 };

    int hists_size = (nbins * CELLS_PER_BLOCK_X * CELLS_PER_BLOCK_Y * 12) * sizeof(float);
    int final_hists_size = (nbins * CELLS_PER_BLOCK_X * CELLS_PER_BLOCK_Y) * sizeof(float);

    int smem = (hists_size + final_hists_size) * blocks_in_group;

    int idx = 0;
    idx = k.set(idx, block_stride_x);
    idx = k.set(idx, block_stride_y);
    idx = k.set(idx, nbins);
    idx = k.set(idx, (int)block_hist_size);
    idx = k.set(idx, img_block_width);
    idx = k.set(idx, blocks_in_group);
    idx = k.set(idx, blocks_total);
    idx = k.set(idx, grad_quadstep);
    idx = k.set(idx, qangle_step);
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(grad));
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(qangle));
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(gauss_w_lut));
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(block_hists));
    idx = k.set(idx, (void*)NULL, (size_t)smem);

    return k.run(2, globalThreads, localThreads, false);
}

static int power_2up(unsigned int n)
{
    for(unsigned int i = 1; i<=1024; i<<=1)
        if(n < i)
            return i;
    return -1; // Input is too big
}

static bool ocl_normalize_hists(int nbins, int block_stride_x, int block_stride_y,
                                int height, int width, UMat block_hists, float threshold)
{
    int block_hist_size = nbins * CELLS_PER_BLOCK_X * CELLS_PER_BLOCK_Y;
    int img_block_width = (width - CELLS_PER_BLOCK_X * CELL_WIDTH + block_stride_x)
        / block_stride_x;
    int img_block_height = (height - CELLS_PER_BLOCK_Y * CELL_HEIGHT + block_stride_y)
        / block_stride_y;
    int nthreads;
    size_t globalThreads[3] = { 1, 1, 1  };
    size_t localThreads[3] = { 1, 1, 1  };

    int idx = 0;
    bool is_cpu = cv::ocl::Device::getDefault().type() == cv::ocl::Device::TYPE_CPU;
    cv::String opts;
    ocl::Kernel k;
    if ( nbins == 9 )
    {
        k.create("normalize_hists_36_kernel", ocl::objdetect::objdetect_hog_oclsrc, "");
        if(k.empty())
            return false;
        if(is_cpu)
           opts = "-D CPU ";
        else
            opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
        k.create("normalize_hists_36_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
        if(k.empty())
            return false;

        int blocks_in_group = NTHREADS / block_hist_size;
        nthreads = blocks_in_group * block_hist_size;
        int num_groups = (img_block_width * img_block_height + blocks_in_group - 1)/blocks_in_group;
        globalThreads[0] = nthreads * num_groups;
        localThreads[0] = nthreads;
    }
    else
    {
        k.create("normalize_hists_kernel", ocl::objdetect::objdetect_hog_oclsrc, "-D WAVE_SIZE=32");
        if(k.empty())
            return false;
        if(is_cpu)
           opts = "-D CPU ";
        else
            opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
        k.create("normalize_hists_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
        if(k.empty())
            return false;

        nthreads = power_2up(block_hist_size);
        globalThreads[0] = img_block_width * nthreads;
        globalThreads[1] = img_block_height;
        localThreads[0] = nthreads;

        if ((nthreads < 32) || (nthreads > 512) )
            return false;

        idx = k.set(idx, nthreads);
        idx = k.set(idx, block_hist_size);
        idx = k.set(idx, img_block_width);
    }
    idx = k.set(idx, ocl::KernelArg::PtrReadWrite(block_hists));
    idx = k.set(idx, threshold);
    idx = k.set(idx, (void*)NULL,  nthreads * sizeof(float));

    return k.run(2, globalThreads, localThreads, false);
}

static bool ocl_extract_descrs_by_rows(int win_height, int win_width, int block_stride_y, int block_stride_x, int win_stride_y, int win_stride_x,
                                       int height, int width, UMat block_hists, UMat descriptors,
                                       int block_hist_size, int descr_size, int descr_width)
{
    ocl::Kernel k("extract_descrs_by_rows_kernel", ocl::objdetect::objdetect_hog_oclsrc);
    if(k.empty())
        return false;

    int win_block_stride_x = win_stride_x / block_stride_x;
    int win_block_stride_y = win_stride_y / block_stride_y;
    int img_win_width = (width - win_width + win_stride_x) / win_stride_x;
    int img_win_height = (height - win_height + win_stride_y) / win_stride_y;
    int img_block_width = (width - CELLS_PER_BLOCK_X * CELL_WIDTH + block_stride_x) /
        block_stride_x;

    int descriptors_quadstep = (int)descriptors.step >> 2;

    size_t globalThreads[3] = { (size_t)img_win_width * NTHREADS, (size_t)img_win_height, 1 };
    size_t localThreads[3] = { NTHREADS, 1, 1 };

    int idx = 0;
    idx = k.set(idx, block_hist_size);
    idx = k.set(idx, descriptors_quadstep);
    idx = k.set(idx, descr_size);
    idx = k.set(idx, descr_width);
    idx = k.set(idx, img_block_width);
    idx = k.set(idx, win_block_stride_x);
    idx = k.set(idx, win_block_stride_y);
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(block_hists));
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(descriptors));

    return k.run(2, globalThreads, localThreads, false);
}

static bool ocl_extract_descrs_by_cols(int win_height, int win_width, int block_stride_y, int block_stride_x, int win_stride_y, int win_stride_x,
                                       int height, int width, UMat block_hists, UMat descriptors,
                                       int block_hist_size, int descr_size, int nblocks_win_x, int nblocks_win_y)
{
    ocl::Kernel k("extract_descrs_by_cols_kernel", ocl::objdetect::objdetect_hog_oclsrc);
    if(k.empty())
        return false;

    int win_block_stride_x = win_stride_x / block_stride_x;
    int win_block_stride_y = win_stride_y / block_stride_y;
    int img_win_width = (width - win_width + win_stride_x) / win_stride_x;
    int img_win_height = (height - win_height + win_stride_y) / win_stride_y;
    int img_block_width = (width - CELLS_PER_BLOCK_X * CELL_WIDTH + block_stride_x) /
        block_stride_x;

    int descriptors_quadstep = (int)descriptors.step >> 2;

    size_t globalThreads[3] = { (size_t)img_win_width * NTHREADS, (size_t)img_win_height, 1 };
    size_t localThreads[3] = { NTHREADS, 1, 1 };

    int idx = 0;
    idx = k.set(idx, block_hist_size);
    idx = k.set(idx, descriptors_quadstep);
    idx = k.set(idx, descr_size);
    idx = k.set(idx, nblocks_win_x);
    idx = k.set(idx, nblocks_win_y);
    idx = k.set(idx, img_block_width);
    idx = k.set(idx, win_block_stride_x);
    idx = k.set(idx, win_block_stride_y);
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(block_hists));
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(descriptors));

    return k.run(2, globalThreads, localThreads, false);
}

static bool ocl_compute(InputArray _img, Size win_stride, std::vector<float>& _descriptors, HOGDescriptor::DescriptorStorageFormat descr_format, Size blockSize,
                        Size cellSize, int nbins, Size blockStride, Size winSize, float sigma, bool gammaCorrection, double L2HysThreshold, bool signedGradient)
{
    Size imgSize = _img.size();
    Size effect_size = imgSize;

    UMat grad(imgSize, CV_32FC2);
    int qangle_type = ocl::Device::getDefault().isIntel() ? CV_32SC2 : CV_8UC2;
    UMat qangle(imgSize, qangle_type);

    const size_t block_hist_size = getBlockHistogramSize(blockSize, cellSize, nbins);
    const Size blocks_per_img = numPartsWithin(imgSize, blockSize, blockStride);
    UMat block_hists(1, static_cast<int>(block_hist_size * blocks_per_img.area()) + 256, CV_32F);

    Size wins_per_img = numPartsWithin(imgSize, winSize, win_stride);
    UMat labels(1, wins_per_img.area(), CV_8U);

    float scale = 1.f / (2.f * sigma * sigma);
    Mat gaussian_lut(1, 512, CV_32FC1);
    int idx = 0;
    for(int i=-8; i<8; i++)
        for(int j=-8; j<8; j++)
            gaussian_lut.at<float>(idx++) = std::exp(-(j * j + i * i) * scale);
    for(int i=-8; i<8; i++)
        for(int j=-8; j<8; j++)
            gaussian_lut.at<float>(idx++) = (8.f - fabs(j + 0.5f)) * (8.f - fabs(i + 0.5f)) / 64.f;

    if(!ocl_computeGradient(_img, grad, qangle, nbins, effect_size, gammaCorrection, signedGradient))
        return false;

    UMat gauss_w_lut;
    gaussian_lut.copyTo(gauss_w_lut);
    if(!ocl_compute_hists(nbins, blockStride.width, blockStride.height, effect_size.height,
        effect_size.width, grad, qangle, gauss_w_lut, block_hists, block_hist_size))
        return false;

    if(!ocl_normalize_hists(nbins, blockStride.width, blockStride.height, effect_size.height,
        effect_size.width, block_hists, (float)L2HysThreshold))
        return false;

    Size blocks_per_win = numPartsWithin(winSize, blockSize, blockStride);
    wins_per_img = numPartsWithin(effect_size, winSize, win_stride);

    int descr_size = blocks_per_win.area()*(int)block_hist_size;
    int descr_width = (int)block_hist_size*blocks_per_win.width;

    UMat descriptors(wins_per_img.area(), static_cast<int>(blocks_per_win.area() * block_hist_size), CV_32F);
    switch (descr_format)
    {
    case HOGDescriptor::DESCR_FORMAT_ROW_BY_ROW:
        if(!ocl_extract_descrs_by_rows(winSize.height, winSize.width,
            blockStride.height, blockStride.width, win_stride.height, win_stride.width, effect_size.height,
            effect_size.width, block_hists, descriptors, (int)block_hist_size, descr_size, descr_width))
            return false;
        break;
    case HOGDescriptor::DESCR_FORMAT_COL_BY_COL:
        if(!ocl_extract_descrs_by_cols(winSize.height, winSize.width,
            blockStride.height, blockStride.width, win_stride.height, win_stride.width, effect_size.height, effect_size.width,
            block_hists, descriptors, (int)block_hist_size, descr_size, blocks_per_win.width, blocks_per_win.height))
            return false;
        break;
    default:
        return false;
    }
    descriptors.reshape(1, (int)descriptors.total()).getMat(ACCESS_READ).copyTo(_descriptors);
    return true;
}
#endif //HAVE_OPENCL

void HOGDescriptor::compute(InputArray _img, std::vector<float>& descriptors,
    Size winStride, Size padding, const std::vector<Point>& locations) const
{
    CV_INSTRUMENT_REGION();

    if( winStride == Size() )
        winStride = cellSize;
    Size cacheStride(gcd(winStride.width, blockStride.width),
                     gcd(winStride.height, blockStride.height));

    Size imgSize = _img.size();

    size_t nwindows = locations.size();
    padding.width = (int)alignSize(std::max(padding.width, 0), cacheStride.width);
    padding.height = (int)alignSize(std::max(padding.height, 0), cacheStride.height);
    Size paddedImgSize(imgSize.width + padding.width*2, imgSize.height + padding.height*2);

    CV_OCL_RUN(_img.dims() <= 2 && _img.type() == CV_8UC1 && _img.isUMat(),
        ocl_compute(_img, winStride, descriptors, HOGDescriptor::DESCR_FORMAT_COL_BY_COL, blockSize,
        cellSize, nbins, blockStride, winSize, (float)getWinSigma(), gammaCorrection, L2HysThreshold, signedGradient))

    Mat img = _img.getMat();
    HOGCache cache(this, img, padding, padding, nwindows == 0, cacheStride);

    if( !nwindows )
        nwindows = cache.windowsInImage(paddedImgSize, winStride).area();

    const HOGCache::BlockData* blockData = &cache.blockData[0];

    int nblocks = cache.nblocks.area();
    int blockHistogramSize = cache.blockHistogramSize;
    size_t dsize = getDescriptorSize();
    descriptors.resize(dsize*nwindows);

    // for each window
    for( size_t i = 0; i < nwindows; i++ )
    {
        float* descriptor = &descriptors[i*dsize];

        Point pt0;
        if( !locations.empty() )
        {
            pt0 = locations[i];
            if( pt0.x < -padding.width || pt0.x > img.cols + padding.width - winSize.width ||
                pt0.y < -padding.height || pt0.y > img.rows + padding.height - winSize.height )
                continue;
        }
        else
        {
            pt0 = cache.getWindow(paddedImgSize, winStride, (int)i).tl() - Point(padding);
//            CV_Assert(pt0.x % cacheStride.width == 0 && pt0.y % cacheStride.height == 0);
        }

        for( int j = 0; j < nblocks; j++ )
        {
            const HOGCache::BlockData& bj = blockData[j];
            Point pt = pt0 + bj.imgOffset;

            float* dst = descriptor + bj.histOfs;
            const float* src = cache.getBlock(pt, dst);
            if( src != dst )
                memcpy(dst, src, blockHistogramSize * sizeof(float));
        }
    }
}

void HOGDescriptor::detect(InputArray _img,
    std::vector<Point>& hits, std::vector<double>& weights, double hitThreshold,
    Size winStride, Size padding, const std::vector<Point>& locations) const
{
    CV_INSTRUMENT_REGION();

    Mat img = _img.getMat();
    hits.clear();
    weights.clear();
    if( svmDetector.empty() )
        return;

    if( winStride == Size() )
        winStride = cellSize;
    Size cacheStride(gcd(winStride.width, blockStride.width),
        gcd(winStride.height, blockStride.height));

    size_t nwindows = locations.size();
    padding.width = (int)alignSize(std::max(padding.width, 0), cacheStride.width);
    padding.height = (int)alignSize(std::max(padding.height, 0), cacheStride.height);
    Size paddedImgSize(img.cols + padding.width*2, img.rows + padding.height*2);

    HOGCache cache(this, img, padding, padding, nwindows == 0, cacheStride);

    if( !nwindows )
        nwindows = cache.windowsInImage(paddedImgSize, winStride).area();

    const HOGCache::BlockData* blockData = &cache.blockData[0];

    int nblocks = cache.nblocks.area();
    int blockHistogramSize = cache.blockHistogramSize;
    size_t dsize = getDescriptorSize();

    double rho = svmDetector.size() > dsize ? svmDetector[dsize] : 0;
    std::vector<float> blockHist(blockHistogramSize);

#if CV_SSE2 || CV_NEON
    float partSum[4];
#endif

    for( size_t i = 0; i < nwindows; i++ )
    {
        Point pt0;
        if( !locations.empty() )
        {
            pt0 = locations[i];
            if( pt0.x < -padding.width || pt0.x > img.cols + padding.width - winSize.width ||
                    pt0.y < -padding.height || pt0.y > img.rows + padding.height - winSize.height )
                continue;
        }
        else
        {
            pt0 = cache.getWindow(paddedImgSize, winStride, (int)i).tl() - Point(padding);
            CV_Assert(pt0.x % cacheStride.width == 0 && pt0.y % cacheStride.height == 0);
        }
        double s = rho;
        const float* svmVec = &svmDetector[0];

        int j, k;
        for( j = 0; j < nblocks; j++, svmVec += blockHistogramSize )
        {
            const HOGCache::BlockData& bj = blockData[j];
            Point pt = pt0 + bj.imgOffset;

            const float* vec = cache.getBlock(pt, &blockHist[0]);
#if CV_SSE2
            __m128 _vec = _mm_loadu_ps(vec);
            __m128 _svmVec = _mm_loadu_ps(svmVec);
            __m128 sum = _mm_mul_ps(_svmVec, _vec);

            for( k = 4; k <= blockHistogramSize - 4; k += 4 )
            {
                _vec = _mm_loadu_ps(vec + k);
                _svmVec = _mm_loadu_ps(svmVec + k);

                sum = _mm_add_ps(sum, _mm_mul_ps(_vec, _svmVec));
            }

            _mm_storeu_ps(partSum, sum);
            double t0 = partSum[0] + partSum[1];
            double t1 = partSum[2] + partSum[3];
            s += t0 + t1;
#elif CV_NEON
            float32x4_t _vec = vld1q_f32(vec);
            float32x4_t _svmVec = vld1q_f32(svmVec);
            float32x4_t sum = vmulq_f32(_svmVec, _vec);

            for( k = 4; k <= blockHistogramSize - 4; k += 4 )
            {
                _vec = vld1q_f32(vec + k);
                _svmVec = vld1q_f32(svmVec + k);

                sum = vaddq_f32(sum, vmulq_f32(_vec, _svmVec));
            }

            vst1q_f32(partSum, sum);
            double t0 = partSum[0] + partSum[1];
            double t1 = partSum[2] + partSum[3];
            s += t0 + t1;
#else
            for( k = 0; k <= blockHistogramSize - 4; k += 4 )
                s += vec[k]*svmVec[k] + vec[k+1]*svmVec[k+1] +
                    vec[k+2]*svmVec[k+2] + vec[k+3]*svmVec[k+3];
#endif
            for( ; k < blockHistogramSize; k++ )
                s += vec[k]*svmVec[k];
        }
        if( s >= hitThreshold )
        {
            hits.push_back(pt0);
            weights.push_back(s);
        }
    }
}

void HOGDescriptor::detect(InputArray img, std::vector<Point>& hits, double hitThreshold,
    Size winStride, Size padding, const std::vector<Point>& locations) const
{
    CV_INSTRUMENT_REGION();

    std::vector<double> weightsV;
    detect(img, hits, weightsV, hitThreshold, winStride, padding, locations);
}

class HOGInvoker :
    public ParallelLoopBody
{
public:
    HOGInvoker( const HOGDescriptor* _hog, const Mat& _img,
        double _hitThreshold, const Size& _winStride, const Size& _padding,
        const double* _levelScale, std::vector<Rect> * _vec, Mutex* _mtx,
        std::vector<double>* _weights=0, std::vector<double>* _scales=0 )
    {
        hog = _hog;
        img = _img;
        hitThreshold = _hitThreshold;
        winStride = _winStride;
        padding = _padding;
        levelScale = _levelScale;
        vec = _vec;
        weights = _weights;
        scales = _scales;
        mtx = _mtx;
    }

    void operator()(const Range& range) const CV_OVERRIDE
    {
        int i, i1 = range.start, i2 = range.end;
        double minScale = i1 > 0 ? levelScale[i1] : i2 > 1 ? levelScale[i1+1] : std::max(img.cols, img.rows);
        Size maxSz(cvCeil(img.cols/minScale), cvCeil(img.rows/minScale));
        Mat smallerImgBuf(maxSz, img.type());
        std::vector<Point> locations;
        std::vector<double> hitsWeights;

        for( i = i1; i < i2; i++ )
        {
            double scale = levelScale[i];
            Size sz(cvRound(img.cols/scale), cvRound(img.rows/scale));
            Mat smallerImg(sz, img.type(), smallerImgBuf.ptr());
            if( sz == img.size() )
                smallerImg = Mat(sz, img.type(), img.data, img.step);
            else
                resize(img, smallerImg, sz, 0, 0, INTER_LINEAR_EXACT);
            hog->detect(smallerImg, locations, hitsWeights, hitThreshold, winStride, padding);
            Size scaledWinSize = Size(cvRound(hog->winSize.width*scale), cvRound(hog->winSize.height*scale));

            mtx->lock();
            for( size_t j = 0; j < locations.size(); j++ )
            {
                vec->push_back(Rect(cvRound(locations[j].x*scale),
                                    cvRound(locations[j].y*scale),
                                    scaledWinSize.width, scaledWinSize.height));
                if (scales)
                    scales->push_back(scale);
            }
            mtx->unlock();

            if (weights && (!hitsWeights.empty()))
            {
                mtx->lock();
                for (size_t j = 0; j < locations.size(); j++)
                    weights->push_back(hitsWeights[j]);
                mtx->unlock();
            }
        }
    }

private:
    const HOGDescriptor* hog;
    Mat img;
    double hitThreshold;
    Size winStride;
    Size padding;
    const double* levelScale;
    std::vector<Rect>* vec;
    std::vector<double>* weights;
    std::vector<double>* scales;
    Mutex* mtx;
};

#ifdef HAVE_OPENCL

static bool ocl_classify_hists(int win_height, int win_width, int block_stride_y, int block_stride_x,
                               int win_stride_y, int win_stride_x, int height, int width,
                               const UMat& block_hists, UMat detector,
                               float free_coef, float threshold, UMat& labels, Size descr_size, int block_hist_size)
{
    int nthreads;
    bool is_cpu = cv::ocl::Device::getDefault().type() == cv::ocl::Device::TYPE_CPU;
    cv::String opts;

    ocl::Kernel k;
    int idx = 0;
    switch (descr_size.width)
    {
    case 180:
        nthreads = 180;
        k.create("classify_hists_180_kernel", ocl::objdetect::objdetect_hog_oclsrc, "-D WAVE_SIZE=32");
        if(k.empty())
            return false;
        if(is_cpu)
           opts = "-D CPU ";
        else
            opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
        k.create("classify_hists_180_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
        if(k.empty())
            return false;
        idx = k.set(idx, descr_size.width);
        idx = k.set(idx, descr_size.height);
        break;

    case 252:
        nthreads = 256;
        k.create("classify_hists_252_kernel", ocl::objdetect::objdetect_hog_oclsrc, "-D WAVE_SIZE=32");
        if(k.empty())
            return false;
        if(is_cpu)
           opts = "-D CPU ";
        else
            opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
        k.create("classify_hists_252_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
        if(k.empty())
            return false;
        idx = k.set(idx, descr_size.width);
        idx = k.set(idx, descr_size.height);
        break;

    default:
        nthreads = 256;
        k.create("classify_hists_kernel", ocl::objdetect::objdetect_hog_oclsrc, "-D WAVE_SIZE=32");
        if(k.empty())
            return false;
        if(is_cpu)
           opts = "-D CPU ";
        else
            opts = cv::format("-D WAVE_SIZE=%zu", k.preferedWorkGroupSizeMultiple());
        k.create("classify_hists_kernel", ocl::objdetect::objdetect_hog_oclsrc, opts);
        if(k.empty())
            return false;
        idx = k.set(idx, descr_size.area());
        idx = k.set(idx, descr_size.height);
    }

    int win_block_stride_x = win_stride_x / block_stride_x;
    int win_block_stride_y = win_stride_y / block_stride_y;
    int img_win_width = (width - win_width + win_stride_x) / win_stride_x;
    int img_win_height = (height - win_height + win_stride_y) / win_stride_y;
    int img_block_width = (width - CELLS_PER_BLOCK_X * CELL_WIDTH + block_stride_x) /
        block_stride_x;

    size_t globalThreads[3] = { (size_t)img_win_width * nthreads, (size_t)img_win_height, 1 };
    size_t localThreads[3] = { (size_t)nthreads, 1, 1 };

    idx = k.set(idx, block_hist_size);
    idx = k.set(idx, img_win_width);
    idx = k.set(idx, img_block_width);
    idx = k.set(idx, win_block_stride_x);
    idx = k.set(idx, win_block_stride_y);
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(block_hists));
    idx = k.set(idx, ocl::KernelArg::PtrReadOnly(detector));
    idx = k.set(idx, free_coef);
    idx = k.set(idx, threshold);
    idx = k.set(idx, ocl::KernelArg::PtrWriteOnly(labels));

    return k.run(2, globalThreads, localThreads, false);
}

static bool ocl_detect(InputArray img, std::vector<Point> &hits, double hit_threshold, Size win_stride,
                       const UMat& oclSvmDetector, Size blockSize, Size cellSize, int nbins, Size blockStride, Size winSize,
                       bool gammaCorrection, double L2HysThreshold, float sigma, float free_coef, bool signedGradient)
{
    hits.clear();
    if (oclSvmDetector.empty())
        return false;

    Size imgSize = img.size();
    Size effect_size = imgSize;
    UMat grad(imgSize, CV_32FC2);
    int qangle_type = ocl::Device::getDefault().isIntel() ? CV_32SC2 : CV_8UC2;
    UMat qangle(imgSize, qangle_type);

    const size_t block_hist_size = getBlockHistogramSize(blockSize, cellSize, nbins);
    const Size blocks_per_img = numPartsWithin(imgSize, blockSize, blockStride);
    UMat block_hists(1, static_cast<int>(block_hist_size * blocks_per_img.area()) + 256, CV_32F);

    Size wins_per_img = numPartsWithin(imgSize, winSize, win_stride);
    UMat labels(1, wins_per_img.area(), CV_8U);

    float scale = 1.f / (2.f * sigma * sigma);
    Mat gaussian_lut(1, 512, CV_32FC1);
    int idx = 0;
    for(int i=-8; i<8; i++)
        for(int j=-8; j<8; j++)
            gaussian_lut.at<float>(idx++) = std::exp(-(j * j + i * i) * scale);
    for(int i=-8; i<8; i++)
        for(int j=-8; j<8; j++)
            gaussian_lut.at<float>(idx++) = (8.f - fabs(j + 0.5f)) * (8.f - fabs(i + 0.5f)) / 64.f;

    if(!ocl_computeGradient(img, grad, qangle, nbins, effect_size, gammaCorrection, signedGradient))
        return false;

    UMat gauss_w_lut;
    gaussian_lut.copyTo(gauss_w_lut);
    if(!ocl_compute_hists(nbins, blockStride.width, blockStride.height, effect_size.height,
        effect_size.width, grad, qangle, gauss_w_lut, block_hists, block_hist_size))
        return false;

    if(!ocl_normalize_hists(nbins, blockStride.width, blockStride.height, effect_size.height,
        effect_size.width, block_hists, (float)L2HysThreshold))
        return false;

    Size blocks_per_win = numPartsWithin(winSize, blockSize, blockStride);

    Size descr_size((int)block_hist_size*blocks_per_win.width, blocks_per_win.height);

    if(!ocl_classify_hists(winSize.height, winSize.width, blockStride.height,
        blockStride.width, win_stride.height, win_stride.width,
        effect_size.height, effect_size.width, block_hists, oclSvmDetector,
        free_coef, (float)hit_threshold, labels, descr_size, (int)block_hist_size))
        return false;

    Mat labels_host = labels.getMat(ACCESS_READ);
    unsigned char *vec = labels_host.ptr();
    for (int i = 0; i < wins_per_img.area(); i++)
    {
        int y = i / wins_per_img.width;
        int x = i - wins_per_img.width * y;
        if (vec[i])
        {
            hits.push_back(Point(x * win_stride.width, y * win_stride.height));
        }
    }
    return true;
}

static bool ocl_detectMultiScale(InputArray _img, std::vector<Rect> &found_locations, std::vector<double>& level_scale,
                                              double hit_threshold, Size win_stride, double group_threshold,
                                              const UMat& oclSvmDetector, Size blockSize, Size cellSize,
                                              int nbins, Size blockStride, Size winSize, bool gammaCorrection,
                                              double L2HysThreshold, float sigma, float free_coef, bool signedGradient)
{
    std::vector<Rect> all_candidates;
    std::vector<Point> locations;
    UMat image_scale;
    Size imgSize = _img.size();
    image_scale.create(imgSize, _img.type());

    for (size_t i = 0; i<level_scale.size() ; i++)
    {
        double scale = level_scale[i];
        Size effect_size = Size(cvRound(imgSize.width / scale), cvRound(imgSize.height / scale));
        if (effect_size == imgSize)
        {
            if(!ocl_detect(_img, locations, hit_threshold, win_stride, oclSvmDetector, blockSize, cellSize, nbins,
                blockStride, winSize, gammaCorrection, L2HysThreshold, sigma, free_coef, signedGradient))
                return false;
        }
        else
        {
            resize(_img, image_scale, effect_size, 0, 0, INTER_LINEAR_EXACT);
            if(!ocl_detect(image_scale, locations, hit_threshold, win_stride, oclSvmDetector, blockSize, cellSize, nbins,
                blockStride, winSize, gammaCorrection, L2HysThreshold, sigma, free_coef, signedGradient))
                return false;
        }
        Size scaled_win_size(cvRound(winSize.width * scale),
            cvRound(winSize.height * scale));
        for (size_t j = 0; j < locations.size(); j++)
            all_candidates.push_back(Rect(Point2d(locations[j]) * scale, scaled_win_size));
    }
    found_locations.assign(all_candidates.begin(), all_candidates.end());
    groupRectangles(found_locations, (int)group_threshold, 0.2);
    clipObjects(imgSize, found_locations, 0, 0);

    return true;
}
#endif //HAVE_OPENCL

void HOGDescriptor::detectMultiScale(
    InputArray _img, std::vector<Rect>& foundLocations, std::vector<double>& foundWeights,
    double hitThreshold, Size winStride, Size padding,
    double scale0, double finalThreshold, bool useMeanshiftGrouping) const
{
    CV_INSTRUMENT_REGION();

    double scale = 1.;
    int levels = 0;

    Size imgSize = _img.size();
    std::vector<double> levelScale;
    for( levels = 0; levels < nlevels; levels++ )
    {
        levelScale.push_back(scale);
        if( cvRound(imgSize.width/scale) < winSize.width ||
            cvRound(imgSize.height/scale) < winSize.height ||
                scale0 <= 1 )
            break;
        scale *= scale0;
    }
    levels = std::max(levels, 1);
    levelScale.resize(levels);

    if(winStride == Size())
        winStride = blockStride;

    CV_OCL_RUN(_img.dims() <= 2 && _img.type() == CV_8UC1 && scale0 > 1 && winStride.width % blockStride.width == 0 &&
        winStride.height % blockStride.height == 0 && padding == Size(0,0) && _img.isUMat(),
        ocl_detectMultiScale(_img, foundLocations, levelScale, hitThreshold, winStride, finalThreshold, oclSvmDetector,
        blockSize, cellSize, nbins, blockStride, winSize, gammaCorrection, L2HysThreshold, (float)getWinSigma(), free_coef, signedGradient));

    std::vector<Rect> allCandidates;
    std::vector<double> tempScales;
    std::vector<double> tempWeights;
    std::vector<double> foundScales;

    Mutex mtx;
    Mat img = _img.getMat();
    Range range(0, (int)levelScale.size());
    HOGInvoker invoker(this, img, hitThreshold, winStride, padding, &levelScale[0], &allCandidates, &mtx, &tempWeights, &tempScales);
    parallel_for_(range, invoker);

    std::copy(tempScales.begin(), tempScales.end(), back_inserter(foundScales));
    foundLocations.clear();
    std::copy(allCandidates.begin(), allCandidates.end(), back_inserter(foundLocations));
    foundWeights.clear();
    std::copy(tempWeights.begin(), tempWeights.end(), back_inserter(foundWeights));

    if ( useMeanshiftGrouping )
        groupRectangles_meanshift(foundLocations, foundWeights, foundScales, finalThreshold, winSize);
    else
        groupRectangles(foundLocations, foundWeights, (int)finalThreshold, 0.2);
    clipObjects(imgSize, foundLocations, 0, &foundWeights);
}

void HOGDescriptor::detectMultiScale(InputArray img, std::vector<Rect>& foundLocations,
    double hitThreshold, Size winStride, Size padding,
    double scale0, double finalThreshold, bool useMeanshiftGrouping) const
{
    CV_INSTRUMENT_REGION();

    std::vector<double> foundWeights;
    detectMultiScale(img, foundLocations, foundWeights, hitThreshold, winStride,
                padding, scale0, finalThreshold, useMeanshiftGrouping);
}

template<typename _ClsName> struct RTTIImpl
{
public:
    static int isInstance(const void* ptr)
    {
        static _ClsName dummy;
        static void* dummyp = &dummy;
        union
        {
            const void* p;
            const void** pp;
        } a, b;
        a.p = dummyp;
        b.p = ptr;
        return *a.pp == *b.pp;
    }
    static void release(void** dbptr)
    {
        if(dbptr && *dbptr)
        {
            delete (_ClsName*)*dbptr;
            *dbptr = 0;
        }
    }
    static void* read(CvFileStorage* fs, CvFileNode* n)
    {
        FileNode fn(fs, n);
        _ClsName* obj = new _ClsName;
        if(obj->read(fn))
            return obj;
        delete obj;
        return 0;
    }

    static void write(CvFileStorage* _fs, const char* name, const void* ptr, CvAttrList)
    {
        if(ptr && _fs)
        {
            FileStorage fs(_fs, false);
            ((const _ClsName*)ptr)->write(fs, String(name));
        }
    }

    static void* clone(const void* ptr)
    {
        if(!ptr)
            return 0;
        return new _ClsName(*(const _ClsName*)ptr);
    }
};

typedef RTTIImpl<HOGDescriptor> HOGRTTI;

CvType hog_type( CV_TYPE_NAME_HOG_DESCRIPTOR, HOGRTTI::isInstance,
    HOGRTTI::release, HOGRTTI::read, HOGRTTI::write, HOGRTTI::clone);

std::vector<float> HOGDescriptor::getDefaultPeopleDetector()
{
    static const float detector[] = {
        0.05359386f, -0.14721455f, -0.05532170f, 0.05077307f,
        0.11547081f, -0.04268804f, 0.04635834f, -0.05468199f, 0.08232084f,
        0.10424068f, -0.02294518f, 0.01108519f, 0.01378693f, 0.11193510f,
        0.01268418f, 0.08528346f, -0.06309239f, 0.13054633f, 0.08100729f,
        -0.05209739f, -0.04315529f, 0.09341384f, 0.11035026f, -0.07596218f,
        -0.05517511f, -0.04465296f, 0.02947334f, 0.04555536f,
        -3.55954492e-003f, 0.07818956f, 0.07730991f, 0.07890715f, 0.06222893f,
        0.09001380f, -0.03574381f, 0.03414327f, 0.05677258f, -0.04773581f,
        0.03746637f, -0.03521175f, 0.06955440f, -0.03849038f, 0.01052293f,
        0.01736112f, 0.10867710f, 0.08748853f, 3.29739624e-003f, 0.10907028f,
        0.07913758f, 0.10393070f, 0.02091867f, 0.11594022f, 0.13182420f,
        0.09879354f, 0.05362710f, -0.06745391f, -7.01260753e-003f,
        5.24702156e-003f, 0.03236255f, 0.01407916f, 0.02207983f, 0.02537322f,
        0.04547948f, 0.07200756f, 0.03129894f, -0.06274468f, 0.02107014f,
        0.06035208f, 0.08636236f, 4.53164103e-003f, 0.02193363f, 0.02309801f,
        0.05568166f, -0.02645093f, 0.04448695f, 0.02837519f, 0.08975694f,
        0.04461516f, 0.08975355f, 0.07514391f, 0.02306982f, 0.10410084f,
        0.06368385f, 0.05943464f, 4.58420580e-003f, 0.05220337f, 0.06675851f,
        0.08358569f, 0.06712101f, 0.06559004f, -0.03930482f, -9.15936660e-003f,
        -0.05897915f, 0.02816453f, 0.05032348f, 0.06780671f, 0.03377650f,
        -6.09417039e-004f, -0.01795146f, -0.03083684f, -0.01302475f,
        -0.02972313f, 7.88706727e-003f, -0.03525961f, -2.50397739e-003f,
        0.05245084f, 0.11791293f, -0.02167498f, 0.05299332f, 0.06640524f,
        0.05190265f, -8.27316567e-003f, 0.03033127f, 0.05842173f,
        -4.01050318e-003f, -6.25105947e-003f, 0.05862958f, -0.02465461f,
        0.05546781f, -0.08228195f, -0.07234028f, 0.04640540f, -0.01308254f,
        -0.02506191f, 0.03100746f, -0.04665651f, -0.04591486f, 0.02949927f,
        0.06035462f, 0.02244646f, -0.01698639f, 0.01040041f, 0.01131170f,
        0.05419579f, -0.02130277f, -0.04321722f, -0.03665198f, 0.01126490f,
        -0.02606488f, -0.02228328f, -0.02255680f, -0.03427236f,
        -7.75165204e-003f, -0.06195229f, 8.21638294e-003f, 0.09535975f,
        -0.03709979f, -0.06942501f, 0.14579427f, -0.05448192f, -0.02055904f,
        0.05747357f, 0.02781788f, -0.07077577f, -0.05178314f, -0.10429011f,
        -0.11235505f, 0.07529039f, -0.07559302f, -0.08786739f, 0.02983843f,
        0.02667585f, 0.01382199f, -0.01797496f, -0.03141199f, -0.02098101f,
        0.09029204f, 0.04955018f, 0.13718739f, 0.11379953f, 1.80019124e-003f,
        -0.04577610f, -1.11108483e-003f, -0.09470536f, -0.11596080f,
        0.04489342f, 0.01784211f, 3.06850672e-003f, 0.10781866f,
        3.36498418e-003f, -0.10842580f, -0.07436839f, -0.10535070f,
        -0.01866805f, 0.16057891f, -5.07316366e-003f, -0.04295658f,
        -5.90488780e-003f, 8.82003549e-003f, -0.01492646f, -0.05029279f,
        -0.12875880f, 8.78831954e-004f, -0.01297184f, -0.07592774f,
        -0.02668831f, -6.93787413e-004f, 0.02406698f, -0.01773298f,
        -0.03855745f, -0.05877856f, 0.03259695f, 0.12826584f, 0.06292590f,
        -4.10733931e-003f, 0.10996531f, 0.01332991f, 0.02088735f, 0.04037504f,
        -0.05210760f, 0.07760046f, 0.06399347f, -0.05751930f, -0.10053057f,
        0.07505023f, -0.02139782f, 0.01796176f, 2.34400877e-003f, -0.04208319f,
        0.07355055f, 0.05093350f, -0.02996780f, -0.02219072f, 0.03355330f,
        0.04418742f, -0.05580705f, -0.05037573f, -0.04548179f, 0.01379514f,
        0.02150671f, -0.02194211f, -0.13682702f, 0.05464972f, 0.01608082f,
        0.05309116f, 0.04701022f, 1.33690401e-003f, 0.07575664f, 0.09625306f,
        8.92647635e-003f, -0.02819123f, 0.10866830f, -0.03439325f,
        -0.07092371f, -0.06004780f, -0.02712298f, -7.07467366e-003f,
        -0.01637020f, 0.01336790f, -0.10313606f, 0.04906582f, -0.05732445f,
        -0.02731079f, 0.01042235f, -0.08340668f, 0.03686501f, 0.06108340f,
        0.01322748f, -0.07809529f, 0.03774724f, -0.03413248f, -0.06096525f,
        -0.04212124f, -0.07982176f, -1.25973229e-003f, -0.03045501f,
        -0.01236493f, -0.06312395f, 0.04789570f, -0.04602066f, 0.08576570f,
        0.02521080f, 0.02988098f, 0.10314583f, 0.07060035f, 0.04520544f,
        -0.04426654f, 0.13146530f, 0.08386490f, 0.02164590f, -2.12280243e-003f,
        -0.03686353f, -0.02074944f, -0.03829959f, -0.01530596f, 0.02689708f,
        0.11867401f, -0.06043470f, -0.02785023f, -0.04775074f, 0.04878745f,
        0.06350956f, 0.03494788f, 0.01467400f, 1.17890188e-003f, 0.04379614f,
        2.03681854e-003f, -0.03958609f, -0.01072688f, 6.43705716e-003f,
        0.02996500f, -0.03418507f, -0.01960307f, -0.01219154f,
        -4.37000440e-003f, -0.02549453f, 0.02646318f, -0.01632513f,
        6.46516960e-003f, -0.01929734f, 4.78711911e-003f, 0.04962371f,
        0.03809111f, 0.07265724f, 0.05758125f, -0.03741554f, 0.01648608f,
        -8.45285598e-003f, 0.03996826f, -0.08185477f, 0.02638875f,
        -0.04026615f, -0.02744674f, -0.04071517f, 1.05096330e-003f,
        -0.04741232f, -0.06733172f, 8.70434940e-003f, -0.02192543f,
        1.35350740e-003f, -0.03056974f, -0.02975521f, -0.02887780f,
        -0.01210713f, -0.04828526f, -0.09066251f, -0.09969629f, -0.03665164f,
        -8.88111943e-004f, -0.06826669f, -0.01866150f, -0.03627640f,
        -0.01408288f, 0.01874239f, -0.02075835f, 0.09145175f, -0.03547291f,
        0.05396780f, 0.04198981f, 0.01301925f, -0.03384354f, -0.12201976f,
        0.06830920f, -0.03715654f, 9.55848210e-003f, 5.05685573e-003f,
        0.05659294f, 3.90764466e-003f, 0.02808490f, -0.05518097f, -0.03711621f,
        -0.02835565f, -0.04420464f, -0.01031947f, 0.01883466f,
        -8.49525444e-003f, -0.09419250f, -0.01269387f, -0.02133371f,
        -0.10190815f, -0.07844430f, 2.43644323e-003f, -4.09610150e-003f,
        0.01202551f, -0.06452291f, -0.10593818f, -0.02464746f, -0.02199699f,
        -0.07401930f, 0.07285886f, 8.87513801e-004f, 9.97662079e-003f,
        8.46779719e-003f, 0.03730333f, -0.02905126f, 0.03573337f, -0.04393689f,
        -0.12014472f, 0.03176554f, -2.76015815e-003f, 0.10824566f, 0.05090732f,
        -3.30179278e-003f, -0.05123822f, 5.04784798e-003f, -0.05664124f,
        -5.99415926e-003f, -0.05341901f, -0.01221393f, 0.01291318f,
        9.91760660e-003f, -7.56987557e-003f, -0.06193124f, -2.24549137e-003f,
        0.01987562f, -0.02018840f, -0.06975540f, -0.06601523f, -0.03349112f,
        -0.08910118f, -0.03371435f, -0.07406893f, -0.02248047f, -0.06159951f,
        2.77751544e-003f, -0.05723337f, -0.04792468f, 0.07518548f,
        2.77279224e-003f, 0.04211938f, 0.03100502f, 0.05278448f, 0.03954679f,
        -0.03006846f, -0.03851741f, -0.02792403f, -0.02875333f, 0.01531280f,
        0.02186953f, -0.01989829f, 2.50679464e-003f, -0.10258728f,
        -0.04785743f, -0.02887216f, 3.85063468e-003f, 0.01112236f,
        8.29218887e-003f, -0.04822981f, -0.04503597f, -0.03713100f,
        -0.06988008f, -0.11002295f, -2.69209221e-003f, 1.85383670e-003f,
        -0.05921049f, -0.06105053f, -0.08458050f, -0.04527602f,
        8.90329306e-004f, -0.05875023f, -2.68602883e-003f, -0.01591195f,
        0.03631859f, 0.05493166f, 0.07300330f, 5.53333294e-003f, 0.06400407f,
        0.01847740f, -5.76280477e-003f, -0.03210877f, 4.25160583e-003f,
        0.01166520f, -1.44864211e-003f, 0.02253744f, -0.03367080f, 0.06983195f,
        -4.22323542e-003f, -8.89401045e-003f, -0.07943393f, 0.05199728f,
        0.06065201f, 0.04133492f, 1.44032843e-003f, -0.09585235f, -0.03964731f,
        0.04232114f, 0.01750465f, -0.04487902f, -7.59733608e-003f, 0.02011171f,
        0.04673622f, 0.09011173f, -0.07869188f, -0.04682482f, -0.05080139f,
        -3.99383716e-003f, -0.05346331f, 0.01085723f, -0.03599333f,
        -0.07097908f, 0.03551549f, 0.02680387f, 0.03471529f, 0.01790393f,
        0.05471273f, 9.62048303e-003f, -0.03180215f, 0.05864431f, 0.02330614f,
        0.01633144f, -0.05616681f, -0.10245429f, -0.08302189f, 0.07291322f,
        -0.01972590f, -0.02619633f, -0.02485327f, -0.04627592f,
        1.48853404e-003f, 0.05514185f, -0.01270860f, -0.01948900f, 0.06373586f,
        0.05002292f, -0.03009798f, 8.76216311e-003f, -0.02474238f,
        -0.05504891f, 1.74034527e-003f, -0.03333667f, 0.01524987f, 0.11663762f,
        -1.32344989e-003f, -0.06608453f, 0.05687166f, -6.89525274e-004f,
        -0.04402352f, 0.09450210f, -0.04222684f, -0.05360983f, 0.01779531f,
        0.02561388f, -0.11075410f, -8.77790991e-003f, -0.01099504f,
        -0.10380266f, 0.03103457f, -0.02105741f, -0.07371717f, 0.05146710f,
        0.10581432f, -0.08617968f, -0.02892107f, 0.01092199f, 0.14551543f,
        -2.24320893e-003f, -0.05818033f, -0.07390742f, 0.05701261f,
        0.12937020f, -0.04986651f, 0.10182415f, 0.05028650f, 0.12515625f,
        0.09175041f, 0.06404983f, 0.01523394f, 0.09460562f, 0.06106631f,
        -0.14266998f, -0.02926703f, 0.02762171f, 0.02164151f,
        -9.58488265e-004f, -0.04231362f, -0.09866509f, 0.04322244f,
        0.05872034f, -0.04838847f, 0.06319253f, 0.02443798f, -0.03606876f,
        9.38737206e-003f, 0.04289991f, -0.01027411f, 0.08156885f, 0.08751175f,
        -0.13191354f, 8.16054735e-003f, -0.01452161f, 0.02952677f, 0.03615945f,
        -2.09128903e-003f, 0.02246693f, 0.09623287f, 0.09412123f, -0.02924758f,
        -0.07815186f, -0.02203079f, -2.02566991e-003f, 0.01094733f,
        -0.01442332f, 0.02838561f, 0.11882371f, 7.28798332e-003f, -0.10345965f,
        0.07561217f, -0.02049661f, 4.44177445e-003f, 0.01609347f, -0.04893158f,
        -0.08758243f, -7.67420698e-003f, 0.08862378f, 0.06098121f, 0.06565887f,
        7.32981879e-003f, 0.03558407f, -0.03874352f, -0.02490055f,
        -0.06771075f, 0.09939223f, -0.01066077f, 0.01382995f, -0.07289080f,
        7.47184316e-003f, 0.10621431f, -0.02878659f, 0.02383525f, -0.03274646f,
        0.02137008f, 0.03837290f, 0.02450992f, -0.04296818f, -0.02895143f,
        0.05327370f, 0.01499020f, 0.04998732f, 0.12938657f, 0.09391870f,
        0.04292390f, -0.03359194f, -0.06809492f, 0.01125796f, 0.17290455f,
        -0.03430733f, -0.06255233f, -0.01813114f, 0.11726857f, -0.06127599f,
        -0.08677909f, -0.03429872f, 0.04684938f, 0.08161420f, 0.03538774f,
        0.01833884f, 0.11321855f, 0.03261845f, -0.04826299f, 0.01752407f,
        -0.01796414f, -0.10464549f, -3.30041884e-003f, 2.29343961e-004f,
        0.01457292f, -0.02132982f, -0.02602923f, -9.87351313e-003f,
        0.04273872f, -0.02103316f, -0.07994065f, 0.02614958f, -0.02111666f,
        -0.06964913f, -0.13453490f, -0.06861878f, -6.09341264e-003f,
        0.08251446f, 0.15612499f, 2.46531400e-003f, 8.88424646e-003f,
        -0.04152999f, 0.02054853f, 0.05277953f, -0.03087788f, 0.02817579f,
        0.13939077f, 0.07641046f, -0.03627627f, -0.03015098f, -0.04041540f,
        -0.01360690f, -0.06227205f, -0.02738223f, 0.13577610f, 0.15235767f,
        -0.05392922f, -0.11175954f, 0.02157129f, 0.01146481f, -0.05264937f,
        -0.06595174f, -0.02749175f, 0.11812254f, 0.17404149f, -0.06137035f,
        -0.11003478f, -0.01351621f, -0.01745916f, -0.08577441f, -0.04469909f,
        -0.06106115f, 0.10559758f, 0.20806813f, -0.09174948f, 7.09621934e-004f,
        0.03579374f, 0.07215115f, 0.02221742f, 0.01827742f, -7.90785067e-003f,
        0.01489554f, 0.14519960f, -0.06425831f, 0.02990399f, -1.80181325e-003f,
        -0.01401528f, -0.04171134f, -3.70530109e-003f, -0.09090481f,
        0.09520713f, 0.08845516f, -0.02651753f, -0.03016730f, 0.02562448f,
        0.03563816f, -0.03817881f, 0.01433385f, 0.02256983f, 0.02872120f,
        0.01001934f, -0.06332260f, 0.04338406f, 0.07001807f, -0.04705722f,
        -0.07318907f, 0.02630457f, 0.03106382f, 0.06648342f, 0.10913180f,
        -0.01630815f, 0.02910308f, 0.02895109f, 0.08040254f, 0.06969310f,
        0.06797734f, 6.08639978e-003f, 4.16588830e-003f, 0.08926726f,
        -0.03123648f, 0.02700146f, 0.01168734f, -0.01631594f, 4.61015804e-003f,
        8.51359498e-003f, -0.03544224f, 0.03571994f, 4.29766066e-003f,
        -0.01970077f, -8.79793242e-003f, 0.09607988f, 0.01544222f,
        -0.03923707f, 0.07308586f, 0.06061262f, 1.31683104e-004f,
        -7.98222050e-003f, 0.02399261f, -0.06084389f, -0.02743429f,
        -0.05475523f, -0.04131311f, 0.03559756f, 0.03055342f, 0.02981433f,
        0.14860515f, 0.01766787f, 0.02945257f, 0.04898238f, 0.01026922f,
        0.02811658f, 0.08267091f, 0.02732154f, -0.01237693f, 0.11760156f,
        0.03802063f, -0.03309754f, 5.24957618e-003f, -0.02460510f, 0.02691451f,
        0.05399988f, -0.10133506f, 0.06385437f, -0.01818005f, 0.02259503f,
        0.03573135f, 0.01042848f, -0.04153402f, -0.04043029f, 0.01643575f,
        0.08326677f, 4.61383024e-004f, -0.05308095f, -0.08536223f,
        -1.61011645e-003f, -0.02163720f, -0.01783352f, 0.03859637f,
        0.08498885f, -0.01725216f, 0.08625131f, 0.10995087f, 0.09177644f,
        0.08498347f, 0.07646490f, 0.05580502f, 0.02693516f, 0.09996913f,
        0.09070327f, 0.06667200f, 0.05873008f, -0.02247842f, 0.07772321f,
        0.12408436f, 0.12629253f, -8.41997913e-004f, 0.01477783f, 0.09165990f,
        -2.98401713e-003f, -0.06466447f, -0.07057302f, 2.09516948e-004f,
        0.02210209f, -0.02158809f, -0.08602506f, -0.02284836f,
        4.01876355e-003f, 9.56660323e-003f, -0.02073978f, -0.04635138f,
        -7.59423291e-003f, -0.01377393f, -0.04559359f, -0.13284740f,
        -0.08671406f, -0.03654395f, 0.01142869f, 0.03287891f, -0.04392983f,
        0.06142959f, 0.17710890f, 0.10385257f, 0.01329137f, 0.10067633f,
        0.12450829f, -0.04476709f, 0.09049144f, 0.04589312f, 0.11167907f,
        0.08587538f, 0.04767583f, 1.67188141e-003f, 0.02359802f, -0.03808852f,
        0.03126272f, -0.01919029f, -0.05698918f, -0.02365112f, -0.06519032f,
        -0.05599358f, -0.07097308f, -0.03301812f, -0.04719102f, -0.02566297f,
        0.01324074f, -0.09230672f, -0.05518232f, -0.04712864f, -0.03380903f,
        -0.06719479f, 0.01183908f, -0.09326738f, 0.01642865f, 0.03789867f,
        -6.61567831e-003f, 0.07796386f, 0.07246574f, 0.04706347f, -0.02523437f,
        -0.01696830f, -0.08068866f, 0.06030888f, 0.10527060f, -0.06611756f,
        0.02977346f, 0.02621830f, 0.01913855f, -0.08479366f, -0.06322418f,
        -0.13570616f, -0.07644490f, 9.31900274e-003f, -0.08095149f,
        -0.10197903f, -0.05204025f, 0.01413151f, -0.07800411f, -0.01885122f,
        -0.07509381f, -0.10136326f, -0.05212355f, -0.09944065f,
        -1.33606605e-003f, -0.06342617f, -0.04178550f, -0.12373723f,
        -0.02832736f, -0.06057501f, 0.05830070f, 0.07604282f, -0.06462587f,
        8.02447461e-003f, 0.11580125f, 0.12332212f, 0.01978462f,
        -2.72378162e-003f, 0.05850752f, -0.04674481f, 0.05148062f,
        -2.62542837e-003f, 0.11253355f, 0.09893716f, 0.09785093f, -0.04659257f,
        -0.01102429f, -0.07002308f, 0.03088913f, -0.02565549f, -0.07671449f,
        3.17443861e-003f, -0.10783514f, -0.02314270f, -0.11089555f,
        -0.01024768f, 0.03116021f, -0.04964825f, 0.02281825f, 5.50005678e-003f,
        -0.08427856f, -0.14685495f, -0.07719755f, -0.13342668f, -0.04525511f,
        -0.09914210f, 0.02588859f, 0.03469279f, 0.04664020f, 0.11688190f,
        0.09647275f, 0.10857815f, -0.01448726f, 0.04299758f, -0.06763151f,
        1.33257592e-003f, 0.14331576f, 0.07574340f, 0.09166205f, 0.05674926f,
        0.11325553f, -0.01106494f, 0.02062161f, -0.11484840f, -0.07492137f,
        -0.02864293f, -0.01275638f, -0.06946032f, -0.10101652f, -0.04113498f,
        -0.02214783f, -0.01273942f, -0.07480393f, -0.10556041f, -0.07622112f,
        -0.09988393f, -0.11453961f, -0.12073903f, -0.09412795f, -0.07146588f,
        -0.04054537f, -0.06127083f, 0.04221122f, 0.07688113f, 0.04099256f,
        0.12663734f, 0.14683802f, 0.21761774f, 0.12525328f, 0.18431792f,
        -1.66402373e-003f, 2.37777247e-003f, 0.01445475f, 0.03509416f,
        0.02654697f, 0.01716739f, 0.05374011f, 0.02944174f, 0.11323927f,
        -0.01485456f, -0.01611330f, -1.85554172e-003f, -0.01708549f,
        -0.05435753f, -0.05302101f, 0.05260378f, -0.03582945f,
        -3.42867890e-004f, 1.36076682e-003f, -0.04436073f, -0.04228432f,
        0.03281291f, -0.05480836f, -0.10197772f, -0.07206279f, -0.10741059f,
        -0.02366946f, 0.10278475f, -2.74783419e-003f, -0.03242477f,
        0.02308955f, 0.02835869f, 0.10348799f, 0.19580358f, 0.10252027f,
        0.08039929f, 0.05525554f, -0.13250865f, -0.14395352f, 3.13586881e-003f,
        -0.03387071f, 8.94669443e-003f, 0.05406157f, -4.97324532e-003f,
        -0.01189114f, 2.82919413e-004f, -0.03901557f, -0.04898705f,
        0.02164520f, -0.01382906f, -0.01850416f, 0.01869347f, -0.02450060f,
        0.02291678f, 0.08196463f, 0.03309153f, -0.10629974f, 0.02473924f,
        0.05344394f, -0.02404823f, -0.03243643f, -5.55244600e-003f,
        -0.08009996f, 0.02811539f, 0.04235742f, 0.01859004f, 0.04902123f,
        -0.01438252f, -0.01526853f, 0.02044195f, -0.05008660f, 0.04244113f,
        0.07611816f, 0.04950470f, -0.06020549f, -4.26026015e-003f, 0.13133512f,
        -0.01438738f, -0.01958807f, -0.04044152f, -0.12425045f,
        2.84353318e-003f, -0.05042776f, -0.09121484f, 7.34345755e-003f,
        0.09388847f, 0.11800314f, 4.72295098e-003f, 4.44378285e-003f,
        -0.07984917f, -0.03613737f, 0.04490915f, -0.02246483f, 0.04681071f,
        0.05240871f, 0.02157206f, -0.04603431f, -0.01197929f, -0.02748779f,
        0.13621049f, 0.08812155f, -0.07802048f, 4.86458559e-003f, -0.01598836f,
        0.01024450f, -0.03463517f, -0.02304239f, -0.08692665f, 0.06655128f,
        0.05785803f, -0.12640759f, 0.02307472f, 0.07337402f, 0.07525434f,
        0.04943763f, -0.02241034f, -0.09978238f, 0.14487994f, -0.06570521f,
        -0.07855482f, 0.02830222f, -5.29603509e-004f, -0.04669895f,
        -0.11822784f, -0.12246452f, -0.15365660f, -0.02969127f, 0.08078201f,
        0.13512598f, 0.11505685f, 0.04740673f, 0.01376022f, -0.05852978f,
        -0.01537809f, -0.05541119f, 0.02491065f, -0.02870786f, 0.02760978f,
        0.23836176f, 0.22347429f, 0.10306466f, -0.06919070f, -0.10132039f,
        -0.20198342f, -0.05040560f, 0.27163076f, 0.36987007f, 0.34540465f,
        0.29095781f, 0.05649706f, 0.04125737f, 0.07505883f, -0.02737836f,
        -8.43431335e-003f, 0.07368195f, 0.01653876f, -0.09402955f,
        -0.09574359f, 0.01474337f, -0.07128561f, -0.03460737f, 0.11438941f,
        0.13752601f, -0.06385452f, -0.06310338f, 8.19548313e-003f, 0.11622470f,
        5.05133113e-003f, -0.07602754f, 0.06695660f, 0.25723928f, 0.09037900f,
        0.28826267f, 0.13165380f, -0.05312614f, -0.02137198f, -0.03442232f,
        -0.06255679f, 0.03899667f, 0.18391028f, 0.26016650f, 0.03374462f,
        0.01860465f, 0.19077586f, 0.18160543f, 3.43634398e-003f, -0.03036782f,
        0.19683038f, 0.35378191f, 0.24968483f, -0.03222649f, 0.28972381f,
        0.43091634f, 0.30778357f, 0.02335266f, -0.09877399f, -6.85245218e-003f,
        0.08945240f, -0.08150686f, 0.02792493f, 0.24806842f, 0.17338486f,
        0.06231801f, -0.10432383f, -0.16653322f, -0.13197899f, -0.08531576f,
        -0.19271527f, -0.13536365f, 0.22240199f, 0.39219588f, 0.26597717f,
        -0.01231649f, 0.01016179f, 0.13379875f, 0.12018334f, -0.04852953f,
        -0.07915270f, 0.07036012f, 3.87723115e-003f, -0.06126805f,
        -0.15015170f, -0.11406515f, -0.08556531f, -0.07429333f, -0.16115491f,
        0.13214062f, 0.25691369f, 0.05697750f, 0.06861912f, -6.02903729e-003f,
        -7.94562511e-003f, 0.04799571f, 0.06695165f, -0.01926842f, 0.06206308f,
        0.13450983f, -0.06381495f, -2.98370165e-003f, -0.03482971f,
        7.53991678e-003f, 0.03895611f, 0.11464261f, 0.01669971f,
        8.27818643e-003f, -7.49160210e-003f, -0.11712562f, -0.10650621f,
        -0.10353880f, -0.04994106f, -7.65618810e-004f, 0.03023767f,
        -0.04759270f, -0.07302686f, -0.05825012f, -0.13156348f, -0.10639747f,
        -0.19393684f, -0.09973683f, -0.07918908f, 4.63177625e-004f,
        -6.61382044e-004f, 0.15853868f, 0.08561199f, -0.07660093f,
        -0.08015265f, -0.06164073f, 0.01882577f, -7.29908410e-004f,
        0.06840892f, 0.03843764f, 0.20274927f, 0.22028814f, -5.26101235e-003f,
        0.01452435f, -0.06331623f, 0.02865064f, 0.05673740f, 0.12171564f,
        0.03837196f, 0.03555467f, -0.02662914f, -0.10280123f, -0.06526285f,
        -0.11066351f, -0.08988424f, -0.10103678f, 8.10526591e-003f,
        5.95238712e-003f, 0.02617721f, -0.01705742f, -0.10897956f,
        -0.08004991f, -0.11271993f, -0.06185647f, -0.06103712f, 0.01597041f,
        -0.05923606f, 0.09410726f, 0.22858568f, 0.03263380f, 0.06772990f,
        -0.09003516f, 0.01017870f, 0.01931688f, 0.08628357f, -0.01430009f,
        0.10954945f, 0.16612452f, -0.02434544f, -0.03310068f, -0.04236627f,
        0.01212392f, -6.15046406e-003f, 0.06954194f, 0.03015283f, 0.01787957f,
        0.02781667f, -0.05561153f, -8.96244217e-003f, -0.04971489f,
        0.07510284f, 0.01775282f, 0.05889897f, -0.07981427f, 0.03647643f,
        -3.73833324e-003f, -0.08894575f, -0.06429435f, -0.08068276f,
        0.03567704f, -0.07131936f, -7.21910037e-003f, -0.09566668f,
        0.17886090f, 0.14911725f, 0.02070032f, -0.05017120f, -0.04992622f,
        0.01570143f, -0.09906903f, 0.06456193f, 0.15329507f, 0.18820767f,
        0.11689861f, -0.01178513f, -0.02225163f, -0.01905318f, 0.10271224f,
        -7.27029052e-003f, 0.11664233f, 0.14796902f, 0.07771893f, 0.02400013f,
        -0.05361797f, -0.01972888f, 0.01376177f, 0.06740040f, -0.06525395f,
        0.05726178f, -0.02404981f, -0.14018567f, -0.02074987f, -0.04621970f,
        -0.04688627f, -0.01842059f, 0.07722727f, -0.04852883f, 0.01529004f,
        -0.19639495f, 0.10817073f, 0.03795860f, -0.09435206f, -0.07984378f,
        -0.03383440f, 0.11081333f, 0.02237366f, 0.12703256f, 0.21613893f,
        0.02918790f, 4.66472283e-003f, -0.10274266f, -0.04854131f,
        -3.46305710e-003f, 0.08652268f, 0.02251546f, 0.09636052f, 0.17180754f,
        -0.09272388f, 4.59174305e-004f, -0.11723048f, -0.12210111f,
        -0.15547538f, 0.07218186f, -0.05297846f, 0.03779940f, 0.05150875f,
        -0.03802310f, 0.03870645f, -0.15250699f, -0.08696499f, -0.02021560f,
        0.04118926f, -0.15177974f, 0.01577647f, 0.10249301f, 7.50041893e-003f,
        0.01721806f, -0.06828983f, -0.02397596f, -0.06598977f, -0.04317593f,
        -0.08064980f, 6.66632550e-003f, 0.03333484f, 0.07093620f, 0.08231064f,
        -0.06577903f, -0.06698844f, -0.06984019f, -0.06508023f, -0.14145090f,
        -0.02393239f, 0.06485303f, 8.83263443e-003f, 0.09251080f, -0.07557579f,
        -0.05067699f, -0.09798748f, -0.06703258f, -0.14056294f, 0.03245994f,
        0.12554143f, 0.01761621f, 0.12980327f, -0.04081950f, -0.11906909f,
        -0.14813015f, -0.08376863f, -0.12200681f, 0.04988137f, 0.05424247f,
        -3.90952639e-003f, 0.03255733f, -0.12717837f, -0.07461493f,
        -0.05703964f, -0.01736189f, -0.08026433f, -0.05433894f, -0.01719359f,
        0.02886275f, 0.01772653f, -0.09163518f, 3.57789593e-003f, -0.10129993f,
        -0.02653764f, -0.08131415f, -0.03847986f, -7.62157550e-004f,
        0.06486648f, 0.19675669f, -0.04919156f, -0.07059129f, -0.04857785f,
        -0.01042383f, -0.08328653f, 0.03660302f, -0.03696846f, 0.04969259f,
        0.08241162f, -0.12514858f, -0.06122676f, -0.03750202f,
        6.52989605e-003f, -0.10247213f, 0.02568346f, 4.51781414e-003f,
        -0.03734229f, -0.01131264f, -0.05412074f, 8.89345480e-004f,
        -0.12388977f, -0.05959237f, -0.12418608f, -0.06151643f, -0.07310260f,
        0.02441575f, 0.07023528f, -0.07548289f, -7.57147965e-004f,
        -0.09061348f, -0.08112976f, -0.06920306f, 9.54394229e-003f,
        -0.01219902f, 1.21273217e-003f, -8.88989680e-003f, -0.08309301f,
        -0.04552661f, -0.10739882f, -0.05691034f, -0.13928030f, 0.09027749f,
        0.15123098f, 0.03175976f, 0.17763577f, 3.29913251e-004f, 0.05151888f,
        -0.09844074f, -0.09475287f, -0.08571247f, 0.16241577f, 0.19336018f,
        8.57454538e-003f, 0.11474732f, -0.01493934f, 0.03352379f, -0.08966240f,
        -0.02322310f, 0.02663568f, 0.05448750f, -0.03536883f, -0.07210463f,
        -0.06807277f, -0.03121621f, -0.05932408f, -0.17282860f, -0.15873498f,
        -0.04956378f, 0.01603377f, -0.12385946f, 0.13878587f, 0.21468069f,
        0.13510075f, 0.20992437f, 0.08845878f, 0.08104013f, 0.03754176f,
        0.12173114f, 0.11103114f, 0.10643122f, 0.13941477f, 0.11640384f,
        0.14786847f, 0.01218238f, 0.01160753f, 0.03547940f, 0.08794311f,
        -0.01695384f, -0.07692261f, -0.08236158f, 6.79194089e-003f,
        -0.02458403f, 0.13022894f, 0.10953187f, 0.09857773f, 0.04735930f,
        -0.04353498f, -0.15173385f, -0.17904443f, -0.10450364f, -0.13418166f,
        -0.06633098f, -0.03170381f, -0.06839000f, -0.11350126f, -0.06983913f,
        0.19083543f, 0.17604128f, 0.07730632f, 0.10022651f, 0.36428109f,
        0.28291923f, 0.12688625f, 0.15942036f, 0.14064661f, -0.11201853f,
        -0.13969108f, -0.09088077f, -0.14107047f, 0.05117374f,
        -2.63348082e-003f, -0.10794610f, -0.09715455f, -0.05284977f,
        0.01565668f, 0.05031200f, 0.07021113f, -0.02963028f, 0.01766960f,
        0.08333644f, -0.03211382f, 4.90096770e-003f, 0.05186674f, -0.05045737f,
        -0.09624767f, -0.02525997f, 0.06916669f, 0.01213916f, 0.05333899f,
        -0.03443280f, -0.10055527f, -0.06291115f, 5.42851724e-003f,
        -6.30360236e-003f, 0.02270257f, -0.01769792f, 0.03273688f, 0.07746078f,
        7.77099328e-003f, 0.05041346f, 0.01648103f, -0.02321534f, -0.09930186f,
        -0.02293853f, 0.02034990f, -0.08324204f, 0.08510064f, -0.03732836f,
        -0.06465405f, -0.06086946f, 0.13680504f, -0.11469388f, -0.03896406f,
        -0.07142810f, 2.67581246e-003f, -0.03639632f, -0.09849060f,
        -0.11014334f, 0.17489147f, 0.17610909f, -0.16091567f, -0.07248894f,
        0.01567141f, 0.23742996f, 0.07552249f, -0.06270349f, -0.07303379f,
        0.25442186f, 0.16903116f, -0.08168741f, -0.05913896f, -0.03954096f,
        6.81776879e-003f, -0.05615319f, -0.07303037f, -0.12176382f,
        0.12385108f, 0.22084464f, -0.05543206f, -0.03310431f, 0.05731593f,
        0.19481890f, 0.04016430f, -0.06480758f, -0.12353460f, 0.18733442f,
        -0.09631214f, -0.11192076f, 0.12404587f, 0.15671748f, 0.19256128f,
        0.10895617f, 0.03391477f, -0.13032004f, -0.05626907f, -0.09025607f,
        0.23485197f, 0.27812332f, 0.26725492f, 0.07255980f, 0.16565137f,
        0.22388470f, 0.07441066f, -0.21003133f, -0.08075339f, -0.15031935f,
        0.07023834f, 0.10872041f, 0.18156518f, 0.20037253f, 0.13571967f,
        -0.11915682f, -0.11131983f, -0.18878011f, 0.06074620f, 0.20578890f,
        0.12413109f, 0.03930207f, 0.29176015f, 0.29502738f, 0.27856228f,
        -0.01803601f, 0.16646385f, 0.19268319f, 0.01900682f, 0.06026287f,
        2.35868432e-003f, 0.01558199f, 0.02707230f, 0.11383014f, 0.12103992f,
        0.03907350f, 0.04637353f, 0.09020995f, 0.11919726f, -3.63007211e-003f,
        0.02220155f, 0.10336831f, 0.17351882f, 0.12259731f, 0.18983354f,
        0.15736865f, 0.01160725f, -0.01690723f, -9.69582412e-004f, 0.07213813f,
        0.01161613f, 0.17864859f, 0.24486147f, 0.18208991f, 0.20177495f,
        0.05972528f, -8.93934630e-003f, -0.02316955f, 0.14436610f, 0.14114498f,
        0.05520950f, 0.06353590f, -0.19124921f, 0.10174713f, 0.29414919f,
        0.26448128f, 0.09344960f, 0.15284036f, 0.19797507f, 0.11369792f,
        -0.12722753f, -0.21396367f, -0.02008235f, -0.06566695f, -0.01662150f,
        -0.03937003f, 0.04778343f, 0.05017274f, -0.02299062f, -0.20208496f,
        -0.06395898f, 0.13721776f, 0.22544557f, 0.14888357f, 0.08687132f,
        0.27088094f, 0.32206613f, 0.09782200f, -0.18523243f, -0.17232181f,
        -0.01041531f, 0.04008654f, 0.04199702f, -0.08081299f, -0.03755421f,
        -0.04809646f, -0.05222081f, -0.21709201f, -0.06622940f, 0.02945281f,
        -0.04600435f, -0.05256077f, -0.08432942f, 0.02848100f, 0.03490564f,
        8.28621630e-003f, -0.11051246f, -0.11210597f, -0.01998289f,
        -0.05369405f, -0.08869293f, -0.18799506f, -0.05436598f, -0.05011634f,
        -0.05419716f, -0.06151857f, -0.10827805f, 0.04346735f, 0.04016083f,
        0.01520820f, -0.12173316f, -0.04880285f, -0.01101406f, 0.03250847f,
        -0.06009551f, -0.03082932f, -0.02295134f, -0.06856834f, -0.08775249f,
        -0.23793389f, -0.09174541f, -0.05538322f, -0.04321031f, -0.11874759f,
        -0.04221844f, -0.06070468f, 0.01194489f, 0.02608565f, -0.03892140f,
        -0.01643151f, -0.02602034f, -0.01305472f, 0.03920100f, -0.06514261f,
        0.01126918f, -6.27710763e-003f, -0.02720047f, -0.11133634f,
        0.03300330f, 0.02398472f, 0.04079665f, -0.10564448f, 0.05966159f,
        0.01195221f, -0.03179441f, -0.01692590f, -0.06177841f, 0.01841576f,
        -5.51078189e-003f, -0.06821765f, -0.03191888f, -0.09545476f,
        0.03030550f, -0.04896152f, -0.02914624f, -0.13283344f, -0.04783419f,
        6.07836898e-003f, -0.01449538f, -0.13358212f, -0.09687774f,
        -0.02813793f, 0.01213498f, 0.06650011f, -0.02039067f, 0.13356198f,
        0.05986415f, -9.12760664e-003f, -0.18780160f, -0.11992817f,
        -0.06342237f, 0.01229534f, 0.07143231f, 0.10713009f, 0.11085765f,
        0.06569190f, -0.02956399f, -0.16288325f, -0.13993549f, -0.01292515f,
        0.03833013f, 0.09130384f, -0.05086257f, 0.05617329f, -0.03896667f,
        -0.06282311f, -0.11490010f, -0.14264110f, -0.04530499f, 0.01598189f,
        0.09167797f, 0.08663294f, 0.04885277f, -0.05741219f, -0.07565769f,
        -0.17136464f, -0.02619422f, -0.02477579f, 0.02679587f, 0.11621952f,
        0.08788391f, 0.15520640f, 0.04709549f, 0.04504483f, -0.10214074f,
        -0.12293372f, -0.04820546f, -0.05484834f, 0.05473754f, 0.07346445f,
        0.05577277f, -0.08209965f, 0.03462975f, -0.20962234f, -0.09324598f,
        3.79481679e-003f, 0.03617633f, 0.16742408f, 0.07058107f, 0.10204960f,
        -0.06795346f, 3.22807301e-003f, -0.12589309f, -0.17496960f,
        0.02078314f, -0.07694324f, 0.12184640f, 0.08997164f, 0.04793497f,
        -0.11383379f, -0.08046359f, -0.25716835f, -0.08080962f,
        6.80711539e-003f, -0.02930280f, -3.04938294e-003f, -0.11106286f,
        -0.04628860f, -0.07821649f, 7.70127494e-003f, -0.10247706f,
        1.21042714e-003f, 0.20573859f, -0.03241005f, 8.42972286e-003f,
        0.01946464f, -0.01197973f, -0.14579976f, 0.04233614f,
        -4.14096704e-003f, -0.06866436f, -0.02431862f, -0.13529138f,
        1.25891645e-003f, -0.11425111f, -0.04303651f, -0.01694815f,
        0.05720210f, -0.16040207f, 0.02772896f, 0.05498345f, -0.15010567f,
        0.01450866f, 0.02350303f, -0.04301004f, -0.04951802f, 0.21702233f,
        -0.03159155f, -0.01963303f, 0.18232647f, -0.03263875f,
        -2.88476888e-003f, 0.01587562f, -1.94303901e-003f, -0.07789494f,
        0.04674156f, -6.25576358e-003f, 0.08925962f, 0.21353747f, 0.01254677f,
        -0.06999976f, -0.05931328f, -0.01884327f, -0.04306272f, 0.11794136f,
        0.03842728f, -0.03907030f, 0.05636114f, -0.09766009f, -0.02104000f,
        8.72711372e-003f, -0.02736877f, -0.05112274f, 0.16996814f, 0.02955785f,
        0.02094014f, 0.08414304f, -0.03335762f, -0.03617457f, -0.05808248f,
        -0.08872101f, 0.02927705f, 0.27077839f, 0.06075108f, 0.07478261f,
        0.15282831f, -0.03908454f, -0.05101782f, -9.51998029e-003f,
        -0.03272416f, -0.08735625f, 0.07633440f, -0.07185312f, 0.13841286f,
        0.07812646f, -0.12901451f, -0.05488589f, -0.05644578f, -0.03290703f,
        -0.11184757f, 0.03751570f, -0.05978153f, -0.09155276f, 0.05657315f,
        -0.04328186f, -0.03047933f, -0.01413135f, -0.10181040f, -0.01384013f,
        0.20132534f, -0.01536873f, -0.07641169f, 0.05906778f, -0.07833145f,
        -0.01523801f, -0.07502609f, -0.09461885f, -0.15013233f, 0.16050665f,
        0.09021381f, 0.08473236f, 0.03386267f, -0.09147339f, -0.09170618f,
        -0.08498498f, -0.05119187f, -0.10431040f, 0.01041618f, -0.03064913f,
        0.09340212f, 0.06448522f, -0.03881054f, -0.04985436f, -0.14794017f,
        -0.05200112f, -0.02144495f, 0.04000821f, 0.12420804f, -0.01851651f,
        -0.04116732f, -0.11951703f, -0.04879033f, -0.08722515f, -0.08454733f,
        -0.10549165f, 0.11251976f, 0.10766345f, 0.19201984f, 0.06128913f,
        -0.02734615f, -0.08834923f, -0.16999826f, -0.03548348f,
        -5.36092324e-003f, 0.08297954f, 0.07226378f, 0.04194529f, 0.04668673f,
        8.73902347e-003f, 0.06980139f, 0.05652480f, 0.05879445f, 0.02477076f,
        0.02451423f, 0.12433673f, 0.05600227f, 0.06886370f, 0.03863076f,
        0.07459056f, 0.02264139f, 0.01495469f, 0.06344220f, 0.06945208f,
        0.02931899f, 0.11719371f, 0.04527427f, 0.03248192f, 2.08271481e-003f,
        0.02044626f, 0.11403449f, 0.04303892f, 0.06444661f, 0.04959024f,
        0.08174094f, 0.09240247f, 0.04894639f, 0.02252937f, -0.01652530f,
        0.07587013f, 0.06064249f, 0.13954395f, 0.02772832f, 0.07093039f,
        0.08501238f, 0.01701301f, 0.09055722f, 0.33421436f, 0.20163782f,
        0.09821030f, 0.07951369f, 0.08695120f, -0.12757730f, -0.13865978f,
        -0.06610068f, -0.10985506f, 0.03406816f, -0.01116336f, -0.07281768f,
        -0.13525715f, -0.12844718f, 0.08956250f, 0.09171610f, 0.10092317f,
        0.23385370f, 0.34489515f, 0.09901748f, 0.02002922f, 0.12335990f,
        0.07606190f, -0.14899330f, -0.15634622f, -0.06494618f, -0.01760547f,
        0.03404277f, -0.13208845f, -0.12101169f, -0.18294574f, -0.16560709f,
        0.02183887f, -0.02752613f, 0.01813638f, 0.02000757f, 0.01319924f,
        0.08030242f, 0.01220535f, 2.98233377e-003f, -0.01307070f, 0.05970297f,
        -0.05345284f, -0.03381982f, -9.87543724e-003f, -0.06869387f,
        0.03956730f, -0.03108176f, -0.05732809f, 0.02172386f, 0.04159765f,
        2.62783933e-003f, 0.04813229f, 0.09358983f, -8.18389002e-003f,
        0.01724574f, -0.02547474f, -0.04967288f, -0.02390376f, 0.06640504f,
        -0.06306566f, 0.01137518f, 0.05589378f, -0.08237787f, 0.02455001f,
        -0.03059422f, -0.08953978f, 0.06851497f, 0.07190268f, -0.07610799f,
        7.87237938e-003f, -7.85830803e-003f, 0.06006952f, -0.01126728f,
        -2.85743061e-003f, -0.04772895f, 0.01884944f, 0.15005857f,
        -0.06268821f, -0.01989072f, 0.01138399f, 0.08760451f, 0.03879007f,
        -9.66926850e-003f, -0.08012961f, 0.06414555f, -0.01362950f,
        -0.09135523f, 0.01755159f, 0.04459474f, 0.09650917f, 0.05219948f,
        -2.19440833e-003f, -0.07037939f, -0.01599054f, 0.13103317f,
        -0.02492603f, -0.01032540f, -0.02903307f, 0.04489160f, 0.05148086f,
        0.01858173f, -0.02919228f, 0.08299296f, -0.04590359f, -0.15745632f,
        -0.09068198f, -0.02972453f, 0.12985018f, 0.22320485f, 0.24261914f,
        0.03642650f, -0.05506422f, 2.67413049e-003f, -0.03834032f, 0.06449424f,
        0.03834866f, 0.03816991f, 0.25039271f, 0.34212017f, 0.32433882f,
        0.18824573f, -0.08599839f, -0.17599408f, -0.15317015f, -0.09913155f,
        -0.02856072f, -0.05304699f, -1.06437842e-003f, -0.06641813f,
        -0.07509298f, 0.01463361f, -0.07551918f, -0.04510373f,
        -8.44620075e-003f, 0.01772176f, 0.04068235f, 0.20295307f, 0.15719447f,
        0.05712103f, 0.26296997f, 0.14657754f, 0.01547317f, -0.05052776f,
        -0.03881342f, -0.01437883f, -0.04930177f, 0.11719568f, 0.24098417f,
        0.26468599f, 0.31698579f, 0.10103608f, -0.01096375f, -0.01367013f,
        0.17104232f, 0.20065314f, 2.67622480e-003f, -0.01190034f, 0.18301608f,
        0.09459770f, -0.06357619f, -0.06473801f, 0.01377906f, -0.10032775f,
        -0.06388740f, 3.80393048e-003f, 0.06206078f, 0.10349120f, 0.26804337f,
        8.17918684e-003f, -0.02314351f, 9.34422202e-003f, 0.09198381f,
        0.03681326f, -8.77339672e-003f, -0.09662418f, -0.02715708f,
        0.13503517f, 0.08962728f, -6.57071499e-003f, -0.03201199f, 0.28510824f,
        0.32095715f, 0.18512695f, -0.14230858f, -0.14048551f, -0.07181299f,
        -0.08575408f, -0.08661680f, -0.17416079f, 7.54326640e-004f,
        0.05601677f, 0.13585392f, -0.04960437f, -0.07708392f, 0.10676333f,
        -0.04407546f, -0.07209078f, 0.03663663f, 0.28949317f, 0.41127121f,
        0.27431169f, -0.06900328f, -0.21474190f, -0.15578632f, -0.19555484f,
        -0.15209621f, -0.11269179f, 0.07416003f, 0.18991330f, 0.26858172f,
        0.01952259f, 0.01017922f, 0.02159843f, -4.95165400e-003f, -0.04368168f,
        -0.12721671f, -0.06673957f, -0.11275250f, 0.04413409f, 0.05578312f,
        0.03896771f, 0.03566417f, -0.05871816f, -0.07388090f, -0.17965563f,
        -0.08570268f, -0.15273231f, -0.06022318f, -0.06999847f,
        -6.81510568e-003f, 0.06294262f, -6.54901436e-004f, -0.01128654f,
        -0.02289657f, 0.04849290f, 0.04140804f, 0.23681939f, 0.14545733f,
        0.01989965f, 0.12032662f, 3.87463090e-003f, -6.02597650e-003f,
        -0.05919775f, -0.03067224f, -0.07787777f, 0.10834727f, 0.02153730f,
        0.02765649f, 0.03975543f, -0.12182906f, -0.04900113f, -0.09940100f,
        -0.06453611f, -0.13757215f, -0.03721382f, 0.02827376f, -0.04351249f,
        0.01907038f, -0.10284120f, -0.05671160f, -0.10760647f, -0.09624009f,
        -0.09565596f, -0.01303654f, 0.03080539f, 0.01416511f, 0.05846142f,
        -5.42971538e-003f, 0.06221476f, -0.03320325f, -0.06791797f,
        -0.05791342f, 0.12851369f, 0.14990346f, 0.03634374f, 0.14262885f,
        0.04330391f, 0.05032569f, -0.05631914f, 0.01606137f, 0.04387223f,
        0.22344995f, 0.15722635f, -0.04693628f, 0.03006579f, -2.52882647e-003f,
        0.05717621f, -0.07529724f, -0.02848588f, -0.06868757f,
        -4.51729307e-003f, 0.06466042f, -0.05935378f, -0.04704857f,
        -0.07363959f, 0.04843248f, -0.13421375f, -0.09789340f, -0.10255270f,
        0.03509852f, 0.04751543f, -0.03822323f, 0.09740467f, 0.04762916f,
        0.03940146f, -0.08283259f, 0.09552965f, 0.05038739f, 0.21258622f,
        0.09646992f, 0.03241193f, 0.05167701f, 0.04614570f, 0.04330090f,
        -0.02671840f, -0.06259909f, -0.02301898f, 0.18829170f, 0.10522786f,
        0.04313190f, 0.01670948f, -0.08421925f, 0.05911417f, -0.10582602f,
        -0.04855484f, -0.08373898f, 0.07775915f, 0.03723533f, -0.12047344f,
        4.86345543e-003f, -0.10520902f, 0.06571782f, -0.07528137f,
        -0.03245651f, -0.09869066f, -0.02917477f, -0.18293270f, 0.14810945f,
        9.24033765e-003f, -0.04354914f, 0.02266885f, -0.11872729f,
        -0.04016589f, 0.02830229f, 0.22539048f, 0.20565644f, 0.16701797f,
        0.09019924f, 0.01300652f, 0.09760600f, -0.03675831f, -0.01935448f,
        -0.06894835f, 0.08077277f, 0.19047537f, 0.11312226f, 0.04106043f,
        -0.11187182f, 0.04312806f, -0.18548580f, -0.11287174f, -0.08794551f,
        0.02078281f, -0.15295486f, 0.11806386f, -0.01103218f, -0.15971117f,
        0.02153538f, -0.05232147f, -0.10835317f, -0.13910367f, 0.05920752f,
        -0.10122602f, 0.20174250f, 0.09105796f, -0.01881348f, 0.09559010f,
        -0.03725745f, -0.09442931f, -0.09763174f, 0.05854454f, 0.08287182f,
        0.12919849f, 0.08594352f, -2.49806582e-003f, 0.02398440f,
        5.67950122e-003f, -0.06296340f, -0.12993270f, 0.03855852f, 0.05186560f,
        0.10839908f, -0.03380463f, -0.12654832f, -0.05399339f, -0.07456800f,
        -0.04736232f, -0.10164231f, 0.07496139f, 0.08125214f, 0.07656177f,
        -0.04999603f, -0.12823077f, -0.07692395f, -0.11317524f, -0.09118655f,
        -0.05695669f, 0.10477209f, 0.07468581f, 0.01630048f, -8.00961629e-003f,
        -0.06582128f, -0.04019095f, -0.04682907f, -0.01907842f, -0.10997720f,
        0.04911406f, 0.02931030f, 0.04197735f, -0.05773980f, -0.09670641f,
        -0.03594951f, -0.03402121f, -0.07149299f, -0.10566200f, 0.10601286f,
        0.06340689f, -0.01518632f, -5.96402306e-003f, -0.07628012f,
        -3.52779147e-003f, -0.02683854f, -0.10265494f, -0.02680815f,
        0.16338381f, 0.03103515f, 0.02296976f, 0.01624348f, -0.10831620f,
        -0.02314233f, -0.04789969f, -0.05530700f, -0.06461314f, 0.10494506f,
        0.04642856f, -0.07592955f, -0.06197905f, -0.09042154f, -0.01445521f,
        -0.04297818f, -0.11262015f, -0.11430512f, 0.03174541f, -0.03677487f,
        -0.02963996f, -0.06610169f, -0.13292049f, -0.07059067f, -0.08444111f,
        -0.02640536f, -0.07136250f, 0.04559967f, 0.01459980f, 0.17989251f,
        0.04435328f, -0.12464730f, -0.02871115f, -0.10752209f, -0.03393742f,
        -0.03791408f, 0.02548251f, 0.01956050f, 0.19245651f, 0.13963254f,
        -0.05904696f, -0.07424626f, -0.10411884f, 1.54176133e-003f,
        0.01797429f, 0.13025844f, 0.04547642f, -0.05710349f, -0.10697161f,
        -0.13489437f, -0.06515755f, -0.06406886f, -4.08572936e-003f,
        -0.01336483f, 0.04368737f, -0.11259720f, -0.05701635f, -0.06469971f,
        -0.08346602f, -0.04166770f, -0.05795543f, -0.08247511f, -0.05742628f,
        0.08452254f, -0.03350224f, 0.13980860f, 0.13252275f, 0.07589617f,
        0.07539988f, 0.12155797f, 0.19087289f, 0.15050751f, 0.21250245f,
        0.14206800f, 0.01298489f, 0.07450245f, 0.06559097f, 0.01700557f,
        0.04512971f, 0.16950700f, 0.10261577f, 0.16389982f, 0.05505059f,
        -0.03453077f, 0.08622462f, 0.07935954f, 0.03976260f, 0.02036091f,
        3.95744899e-003f, 0.03267065f, 0.15235919f, 0.01297494f, -0.08109194f,
        0.01407558f, 4.40693414e-003f, -0.15157418f, -0.11390478f,
        -0.07487597f, -7.81322457e-003f, -0.02749545f, -0.10181408f,
        0.13755716f, 0.14007211f, 0.13482562f, 0.27517235f, 0.34251109f,
        0.07639657f, 0.07268607f, 0.19823882f, 0.16135791f, -0.04186463f,
        -0.12784107f, -0.09846287f, 0.03169041f, 0.10974082f, -0.15051922f,
        -0.08916726f, -0.07138767f, -0.04153349f, 6.25418453e-003f,
        0.01266654f, 0.10533249f, 0.12749144f, 0.15148053f, 0.01498513f,
        0.06305949f, -0.01247123f, -0.08778401f, -0.08551880f, -0.11955146f,
        -0.08493572f, -0.02901620f, -0.02394859f, -0.13427313f, -0.11053200f,
        -0.14413260f, -0.15203285f, 0.03972760f, -3.72127310e-004f,
        -0.04200919f, 0.06105104f, 0.01904975f, -0.01106191f,
        -7.27445772e-003f, -0.01520341f, 1.10228511e-003f, -0.04949187f,
        -0.08013099f, 5.72071038e-003f, 0.08415454f, -0.06523152f, 0.03664081f,
        -0.02673042f, -0.12066154f, -0.03702074f, 0.06006580f, 0.01628682f,
        -6.17772620e-003f, 0.08192339f, -3.41629819e-003f, 0.02870512f,
        0.05807141f, 0.04959986f, 0.04618251f, -0.04901629f, -0.10579574f,
        0.02274442f, 0.12070961f, 2.23597488e-003f, 0.09831765f, -0.03019848f,
        -0.11181970f, -0.04961075f, 0.02498928f, -0.03714991f, -0.01619653f,
        0.02643486f, -7.62964319e-003f, -0.02882290f, -0.06242594f,
        -0.08439861f, 0.07220893f, 0.07263952f, 0.01561574f, 0.03091968f,
        0.01708712f, -0.03797151f, -3.18561122e-003f, 0.01624021f,
        -0.02828573f, 0.11284444f, -1.32280716e-003f, -0.07784860f,
        -0.07209100f, 0.03372242f, 0.12154529f, 0.02278104f, -0.05275500f,
        -0.01918484f, 0.12989293f, 0.05424401f, 0.02333086f, 0.04029022f,
        0.12392918f, 0.09495489f, 0.09190340f, 0.07935889f, 8.76816828e-003f,
        0.17148446f, -8.51302687e-003f, -0.08011249f, -0.06796283f,
        0.04884845f, 0.01112272f, -0.07835306f, -1.14811445e-003f,
        -0.03440760f, 0.02845243f, 0.07695542f, -0.07069533f, -0.01151784f,
        -8.53884313e-003f, -0.01662786f, -0.04163864f, 0.05400505f,
        0.02859163f, 0.02921852f, 0.05003135f, -6.85718050e-003f, -0.01632611f,
        0.07780217f, 0.04042810f, -0.01216440f, 3.60914599e-003f, -0.06322435f,
        0.09516726f, 0.12877031f, -9.69162490e-003f, 0.01031179f, 0.05180895f,
        -9.34659224e-003f, -0.01644533f, -0.04849347f, -0.04343236f,
        0.10514783f, 0.08046635f, -0.04615205f, -0.03975486f, -0.01485525f,
        0.13096830f, -0.01517950f, -0.06571898f, -0.04016372f, 0.01849786f,
        0.02439670f, 0.08067258f, 1.74824719e-003f, 0.07053747f, 0.08819518f,
        -5.08352555e-003f, -0.06550863f, -0.08266170f, -0.07780605f,
        0.01453450f, -0.08756890f, 0.01096501f, -8.71319138e-003f, 0.10110464f,
        0.02420769f, -0.06708383f, 0.02007811f, 5.93133038e-003f, 0.05398923f,
        0.07538138f, 0.02049227f, 0.02242589f, 0.04011070f, -1.44875818e-003f,
        -4.19115182e-003f, 0.06367654f, 0.02506934f, 0.02434536f, 0.05879405f,
        -8.22952855e-003f, -0.01242441f, 0.04224926f, -0.01754923f,
        0.05958161f, 0.03818886f, -0.01830363f, -0.04308917f, -0.04422197f,
        -0.02432721f, 0.02264866f, 2.03751423e-003f, 0.01197031f, 0.04439203f,
        0.12169247f, 0.03602713f, -0.02599251f, -1.98226492e-003f, 0.02046336f,
        -0.02639058f, -1.91242550e-003f, -0.09334669f, -0.03595153f,
        -9.88179818e-003f, -0.06848445f, -0.04666303f, -0.09955736f,
        -0.04206430f, 0.02609075f, 9.09005292e-003f, -0.07138551f,
        -4.22313227e-004f, 0.01766645f, 0.02756404f, 0.01308276f, 0.04052891f,
        0.02387515f, 0.05337298f, 0.02500631f, -0.04970853f, -0.12467445f,
        0.17604403f, 0.12256411f, -0.07512254f, 8.70451052e-003f, -0.05697548f,
        -0.03626474f, -8.76623299e-003f, -0.01210897f, -0.09451522f,
        0.07490732f, -0.02008001f, -0.02681278f, -0.06463405f, -0.01517507f,
        7.33757764e-003f, 6.07147906e-003f, -0.09316964f, -0.04575328f,
        0.13261597f, 0.15424870f, -0.01655918f, -0.02772390f, -0.05243644f,
        -0.02356456f, -0.02351753f, -0.10211615f, -0.12873036f, 0.14549787f,
        0.12519856f, 4.38762689e-003f, 0.02795992f, 0.05170322f, 0.09223596f,
        0.05890015f, 0.02376701f, -0.02777346f, 0.09506908f, 0.02328936f,
        -0.02319928f, -0.03218696f, -0.01527841f, -0.01016694f, -0.02674719f,
        0.05137179f, 0.01980666f, 0.06544447f, -0.01746171f, 0.01026380f,
        0.01561806f, 7.97004555e-004f, 0.07601810f, 0.01907250f, -0.03083035f,
        -0.05987392f, 0.09242783f, 0.14555025f, 0.01035827f, 0.03092401f,
        -0.09562709f, -0.03802354f, 0.02531144f, 0.03079449f, -0.07100715f,
        0.03330721f, -2.69116857e-003f, 0.03167490f, 0.05744999f, 0.03259895f,
        1.91266940e-003f, 0.03194578f, 0.07389776f, 0.02198060f, 0.07633314f,
        0.03293105f, -0.09103648f, 0.04718142f, 0.06102672f, -0.01003063f,
        5.85481385e-003f, -0.01522574f, 0.02323526f, 0.10584345f,
        4.35879454e-003f, 0.06107873f, 0.05868603f, -0.03115531f, 0.01214679f,
        0.08567052f, 3.93926632e-003f, -0.02521488f, -1.88425183e-003f,
        0.02038053f, -6.26854831e-004f, 0.04897438f, -0.04280585f,
        -0.04819689f, -0.04812867f, -0.01451186f, 0.05101469f,
        -9.01125465e-003f, -0.03333859f, 0.03917955f, 0.04196448f, 0.04292135f,
        0.02809529f, 0.02999715f, 0.04081348f, 9.10039060e-003f, 0.09703232f,
        0.10379741f, 0.02348725f, -4.72756615e-003f, 0.01027325f, 0.10402658f,
        0.12071823f, 0.09817299f, -0.02612033f, 0.03638414f, 0.05896405f,
        0.04865025f, 0.04793910f, -0.03882321f, -0.02962117f, -0.01222268f,
        0.04071597f, 0.01922777f, -0.02287866f, 0.03328381f, 0.01859092f,
        0.09024994f, 0.03804455f, -0.01424510f, 0.01953739f, 0.02509617f,
        -0.03390914f, -0.05663941f, -0.01641979f, 0.05848591f, 0.04639670f,
        0.02092116f, 0.12911791f, 0.19918139f, 0.07739855f, -7.25806039e-003f,
        0.04074838f, 0.03183993f, 1.39251316e-003f, -0.01428625f, 0.01865480f,
        0.08529541f, 0.13547510f, 0.11189661f, 0.03998901f, 0.09575938f,
        -0.02631102f, -0.03458253f, -0.04749985f, -0.06070716f,
        4.71884012e-003f, 0.06445789f, -0.02450038f, -0.05483776f,
        -0.04657237f, -0.02030717f, -0.03480766f, -0.09397731f, -0.06399718f,
        -0.01804585f, 5.62348310e-003f, -6.64811488e-003f, -0.06517869f,
        6.96210237e-003f, -0.01860148f, -0.04245830f, -0.05850367f,
        -3.24417115e-003f, 0.07700698f, 0.11290991f, 0.09923030f, -0.02970599f,
        0.05592411f, 0.04813979f, -0.09811195f, -0.09357996f, -0.03276114f,
        0.05218338f, 0.04141375f, 3.92977800e-003f, -0.05047480f, 0.15960084f,
        0.04612800f, -0.03114098f, -0.04650044f, -0.03249795f, -0.02425641f,
        -0.04311355f, 0.04307659f, -0.09401883f, -0.04742785f, -0.01254499f,
        -0.06598741f, 3.41369561e-003f, -0.05620445f, -7.28127593e-003f,
        -0.05998361f, -0.03274450f, -0.07376868f, 3.19015374e-003f,
        -0.07733069f, 0.05815864f, -0.02471071f, 0.03850617f, 0.13838784f,
        0.15399861f, 0.01731321f, -0.01477586f, 0.10393341f, 0.05159833f,
        -0.01945555f, -0.03427503f, -0.04867341f, 0.09237480f, 0.10732719f,
        0.06071450f, -0.01355071f, 0.01844356f, -0.03480803f, -0.03796671f,
        2.15628621e-004f, -0.05440186f, 0.01889855f, -0.01443413f,
        -0.02607902f, -0.02938001f, 0.02720689f, -0.06228397f, -0.02970936f,
        -0.03426210f, -0.10280876f, -0.06739304f, -0.05227850f, 0.03360292f,
        -0.11278441f, -0.06966180f, -0.13937433f, 9.10932291e-003f,
        2.52020749e-004f, -4.07359656e-003f, 0.12310639f, 0.09343060f,
        0.07302511f, 0.03222093f, 0.07532879f, 0.03792387f, -0.04985180f,
        0.01804602f, 0.02694195f, 0.13481498f, 0.04601225f, 0.04106982f,
        0.08511057f, 0.12314661f, 0.01320830f, 0.05044121f, -5.52943908e-003f,
        -0.08992624f, -0.02249301f, -0.08181777f, 0.06165213f, -0.03256603f,
        -0.01068920f, -0.01323473f, -0.11970232f, -0.04616347f, -0.12088681f,
        -0.06762606f, -0.08676834f, -0.06434575f, 0.01772529f, 0.03469615f,
        -0.10926618f, 0.03013873f, 0.14030397f, 0.16130108f, 0.17985588f,
        0.11281928f, 0.10530639f, 0.08905948f, 0.07733764f, 0.06695238f,
        0.02142088f, 0.06438877f, 0.09794453f, 0.05745072f, 0.02788557f,
        0.02632830f, 0.07985807f, 4.24902979e-003f, 8.47890321e-003f,
        -0.02679466f, -5.28812688e-003f, -0.02162580f, -0.07490715f,
        -0.08251337f, -0.02056576f, -0.01026194f, -1.15492963e-003f,
        -5.75720915e-004f, -0.07210591f, -0.07320981f, -0.04883312f,
        -0.10897151f, -0.07477258f, -0.08867134f, -0.09222437f, -0.10924666f,
        -0.10430276f, 0.07953499f, 0.02767959f, 0.11393359f, 0.18779543f,
        0.03313421f, 0.02143700f, 0.05852016f, -2.12067598e-003f,
        -3.76984011e-003f, 0.02774167f, -0.03124610f, 0.01465141f, 0.01616004f,
        -0.01391913f, -0.04404102f, -0.05444227f, -0.14684731f, -0.15016587f,
        0.04509468f, 1.29563001e-003f, 0.01398350f, 0.05610404f, -0.04868806f,
        -0.04776716f, -8.16873740e-003f, -2.30126386e-003f, -0.02286313f,
        0.11983398f, -0.04703261f, -0.08814441f, -0.07585249f, -0.10799607f,
        -0.03232087f, 0.01509786f, -0.04843464f, -0.03967846f, 0.09589416f,
        0.01352560f, -0.01458119f, 0.01050829f, -0.03038946f, 0.01608388f,
        1.11975556e-003f, -0.01250656f, 2.86211423e-003f, 0.04333691f,
        -0.14603497f, -0.01946543f, -0.02327525f, -0.01973944f, 0.07944400f,
        -0.02224544f, -0.06701808f, 0.03476532f, 0.11505594f, -0.02712801f,
        -0.01665113f, 0.06315716f, -0.08205860f, 0.07431999f, 0.04915778f,
        -0.04468752f, -0.01490402f, 0.07400476f, -0.11650901f, 0.05102430f,
        0.04559118f, -0.05916039f, 0.08840760f, -0.01587902f, -0.14890194f,
        0.07857784f, 0.04710254f, -0.05381983f, -0.07331945f, -0.03604643f,
        0.15611970f, 0.07649943f, -0.05959348f, -0.02776607f, 0.11098688f,
        0.03758875f, -0.04446875f, 0.04933187f, 0.01345535f, 0.06921103f,
        0.07364785f, 0.05518956f, 0.02899585f, 0.09375840f, 0.10518434f,
        -0.04420241f, 0.01915282f, -3.56386811e-003f, 0.14586878f, 0.10286101f,
        -0.04360626f, -0.12723237f, 0.09076386f, 0.11119842f, -0.06035013f,
        0.09674817f, 0.08938243f, 0.07065924f, 0.02603180f, 5.84815582e-003f,
        -0.05922065f, 0.12360309f, 3.59695964e-003f, 2.99844006e-003f,
        0.03697936f, 0.02043072f, 0.04168725f, 0.01025975f, -0.01359980f,
        -0.01600920f, 0.02581056f, 0.02329250f, 2.98100687e-003f, 0.01629762f,
        0.06652115f, 0.05855627f, 0.01237463f, -0.01297135f, 0.01761587f,
        0.05090865f, 0.06549342f, -0.04425945f, 2.43203156e-003f,
        3.07327788e-003f, 0.06678630f, -0.04303836f, 0.01082393f, -0.06476044f,
        0.04077786f, 0.12441979f, 0.08237778f, 0.07424165f, 0.04065890f,
        0.06905543f, 0.09556347f, 0.12724875f, -0.02132082f, 0.08514154f,
        -0.04175328f, -0.02666954f, 0.01897836f, 0.03317382f, 9.45465732e-003f,
        -0.01238974f, -0.04242500f, -0.01419479f, -0.03545213f, -0.02440874f,
        0.08684119f, 0.04212951f, 0.02462858f, -0.01104825f, -5.01706870e-003f,
        0.02968982f, 0.02597476f, -0.01568939f, 0.04514892f, 0.06974549f,
        0.08670278f, 0.06828108f, 0.10238872f, 0.05405957f, 0.06548470f,
        -0.03763957f, 0.01366090f, 0.07069602f, 0.05363748f, 0.04798120f,
        0.11706422f, 0.05466456f, -0.01869259f, 0.06344382f, 0.03106543f,
        0.08432506f, -0.02061096f, 0.03821088f, -6.92190882e-003f,
        6.40467042e-003f, -0.01271779f, 6.89014705e-005f, 0.04541415f,
        -0.01899539f, -0.05020239f, 0.03000903f, 0.01090422f, 4.52452758e-003f,
        0.02573632f, -0.02388454f, -0.04200457f, 1.72783900e-003f,
        -0.05978370f, -0.02720562f, 0.06573715f, 0.01154317f, 0.01265615f,
        0.07375994f, -9.19828378e-003f, -0.04914120f, 0.02124831f, 0.06455322f,
        0.04372910f, -0.03310043f, 0.03605788f, -6.78055827e-003f,
        9.36202332e-003f, 0.01747596f, -0.06406314f, -0.06812935f, 0.08080816f,
        -0.02778088f, 0.02735260f, 0.06393493f, 0.06652229f, 0.05676993f,
        0.08640018f, -7.59188086e-003f, -0.02012847f, -0.04741159f,
        -0.01657069f, -0.01624399f, 0.05547778f, -2.33309763e-003f,
        0.01120033f, 0.06141156f, -0.06285004f, -0.08732341f, -0.09313398f,
        -0.04267832f, 5.57443965e-003f, 0.04809862f, 0.01773641f,
        5.37361018e-003f, 0.14842421f, -0.06298012f, -0.02935147f, 0.11443478f,
        -0.05034208f, 5.65494271e-003f, 0.02076526f, -0.04577984f,
        -0.04735741f, 0.02961071f, -0.09307127f, -0.04417921f, -0.04990027f,
        -0.03940028f, 0.01306016f, 0.06267900f, 0.03758737f, 0.08460117f,
        0.13858789f, 0.04862388f, -0.06319809f, -0.05655516f, 0.01885816f,
        -0.03285607f, 0.03371567f, -0.07040928f, -0.04514049f, 0.01392166f,
        0.08184422f, -0.07230316f, 0.02386871f, 0.02184591f, 0.02605764f,
        -0.01033954f, 9.29878280e-003f, 7.67351175e-003f, 0.15189242f,
        0.02069071f, -0.09738296f, -0.08894105f, -0.07768748f, 0.02332268f,
        -0.01778995f, -0.03258888f, -0.08180822f, -0.08492987f, 0.02290156f,
        -0.11368170f, -0.03554465f, -0.04533844f, -0.02861580f, 0.06782424f,
        0.01113123f, 0.02453644f, 0.12721945f, 0.08084814f, -0.03607795f,
        0.01109122f, 0.04803548f, -0.03489929f, 0.03399536f, -0.05682014f,
        8.59533902e-003f, -4.27904585e-003f, 0.03230887f, -0.01300198f,
        -0.01038137f, -0.07930113f, 8.33097473e-003f, 0.02296994f,
        -0.01306500f, -0.01881626f, 0.04413369f, 0.05729880f, -0.03761553f,
        0.01942326f, 1.64540811e-003f, -0.03811319f, 0.04190650f, -0.14978096f,
        -0.04514487f, 0.01209545f, -5.46460645e-003f, -0.01647195f,
        7.63064111e-003f, -0.07494587f, 0.08415288f, 0.10020141f, -0.01228561f,
        0.06553826f, 0.04554005f, 0.07890417f, 0.03041138f, 0.01752007f,
        0.09208256f, -3.74419295e-004f, 0.10549527f, 0.04686913f, 0.01894833f,
        -0.02651412f, -4.34682379e-003f, 5.44942822e-003f, 0.01444484f,
        0.05882156f, -0.03336544f, 0.04603891f, -0.10432546f, 0.01923928f,
        0.01842845f, -0.01712168f, -0.02222766f, 0.04693324f, -0.06202956f,
        -0.01422159f, 0.08732220f, -0.07706107f, 0.02661049f, -0.04300238f,
        -0.03092422f, -0.03552184f, -0.01886088f, -0.04979934f, 0.03906401f,
        0.04608644f, 0.04966111f, 0.04275464f, -0.04621769f, -0.02653212f,
        8.57011229e-003f, 0.03839684f, 0.05818764f, 0.03880796f,
        -2.76100676e-004f, 0.03076511f, -0.03266929f, -0.05374557f,
        0.04986527f, -9.45429131e-003f, 0.03582499f, -2.64564669e-003f,
        -1.07461517e-003f, 0.02962313f, -0.01483363f, 0.03060869f, 0.02448327f,
        0.01845641f, 0.03282966f, -0.03534438f, -0.01084059f, -0.01119136f,
        -1.85360224e-003f, -5.94652840e-004f, -0.04451817f, 2.98327743e-003f,
        0.06272484f, -0.02152076f, -3.05971340e-003f, -0.05070828f,
        0.01531762f, 0.01282815f, 0.05167150f, 9.46266949e-003f,
        -3.34558333e-003f, 0.11442288f, -0.03906701f, -2.67325155e-003f,
        0.03069184f, -0.01134165f, 0.02949462f, 0.02879886f, 0.03855566f,
        -0.03450781f, 0.09142872f, -0.02156654f, 0.06075062f, -0.06220816f,
        0.01944680f, 6.68372354e-003f, -0.06656796f, 8.70784000e-003f,
        0.03456013f, 0.02434320f, -0.13236357f, -0.04177035f, -0.02069627f,
        0.01068112f, 0.01505432f, -0.07517391f, -3.83571628e-003f,
        -0.06298508f, -0.02881260f, -0.13101046f, -0.07221562f,
        -5.79945277e-003f, -8.57300125e-003f, 0.03782469f, 0.02762164f,
        0.04942456f, -0.02936396f, 0.09597211f, 0.01921411f, 0.06101191f,
        -0.04787507f, -0.01379578f, -7.40224449e-003f, -0.02220136f,
        -0.01313756f, 7.77558051e-003f, 0.12296968f, 0.02939998f, 0.03594062f,
        -0.07788624f, -0.01133144f, 3.99316690e-004f, -0.06090347f,
        -0.01122066f, -4.68682544e-003f, 0.07633100f, -0.06748922f,
        -0.05640298f, -0.05265681f, -0.01139122f, -0.01624347f, -0.04715714f,
        -0.01099092f, 0.01048561f, 3.28499987e-003f, -0.05810167f,
        -0.07699911f, -0.03330683f, 0.04185145f, 0.03478536f, 0.02275165f,
        0.02304766f, 6.66040834e-003f, 0.10968148f, -5.93013782e-003f,
        -0.04858336f, -0.04203213f, -0.09316786f, -6.13074889e-003f,
        -0.02544625f, 0.01366201f, 9.18555818e-003f, -0.01846578f,
        -0.05622401f, -0.03989377f, -0.07810296f, 6.91275718e-003f,
        0.05957597f, -0.03901334f, 0.01572002f, -0.01193903f,
        -6.89400872e-003f, -0.03093356f, -0.04136098f, -0.01562869f,
        -0.04604580f, 0.02865234f, -0.08678447f, -0.03232484f, -0.05364593f,
        -0.01445016f, -0.07003860f, -0.08669746f, -0.04520775f, 0.04274122f,
        0.03117515f, 0.08175703f, 0.01081109f, 0.06379741f, 0.06199206f,
        0.02865988f, 0.02360346f, 0.06725410f, -0.03248780f, -9.37702879e-003f,
        0.08265898f, -0.02245839f, 0.05125763f, -0.01862395f, 0.01973453f,
        -0.01994494f, -0.10770868f, 0.03180375f, 3.23935156e-003f,
        -0.02142080f, -0.04256190f, 0.04760900f, 0.04282863f, 0.05635953f,
        -0.01870849f, 0.05540622f, -0.03042666f, 0.01455277f, -0.06630179f,
        -0.05843807f, -0.03739681f, -0.09739155f, -0.03220233f, -0.05620182f,
        -0.10381401f, 0.07400211f, 4.20676917e-003f, 0.03258535f,
        2.14308966e-003f, 0.05121966f, -0.01274337f, 0.02384761f, 0.06335578f,
        -0.07905591f, 0.08375625f, -0.07898903f, -0.06508528f, -0.02498444f,
        0.06535810f, 0.03970535f, 0.04895468f, -0.01169566f, -0.03980601f,
        0.05682293f, 0.05925463f, -0.01165808f, -0.07936699f, -0.04208954f,
        0.01333987f, 0.09051196f, 0.10098671f, -0.03974256f, 0.01238771f,
        -0.07501741f, -0.03655440f, -0.04301528f, 0.09216860f,
        4.63579083e-004f, 0.02851115f, 0.02142735f, 1.28244064e-004f,
        0.02879687f, -0.08554889f, -0.04838862f, 0.08135369f, -0.05756533f,
        0.01413900f, 0.03451880f, -0.06619488f, -0.03053130f, 0.02961676f,
        -0.07384635f, 0.01135692f, 0.05283910f, -0.07778034f, -0.02107482f,
        -0.05511716f, -0.13473752f, 0.03030157f, 0.06722020f, -0.06218817f,
        -0.05826827f, 0.06254654f, 0.02895772f, -0.01664000f, -0.03620280f,
        -0.01612278f, -1.46097376e-003f, 0.14013411f, -8.96181818e-003f,
        -0.03250246f, 3.38630192e-003f, 2.64779478e-003f, 0.03359732f,
        -0.02411991f, -0.04229729f, 0.10666174f, -6.66579151f };
    return std::vector<float>(detector, detector + sizeof(detector)/sizeof(detector[0]));
}

// This function renurn 1981 SVM coeffs obtained from daimler's base.
// To use these coeffs the detection window size should be (48,96)
std::vector<float> HOGDescriptor::getDaimlerPeopleDetector()
{
    static const float detector[] = {
        0.294350f, -0.098796f, -0.129522f, 0.078753f,
        0.387527f, 0.261529f, 0.145939f, 0.061520f,
        0.328699f, 0.227148f, -0.066467f, -0.086723f,
        0.047559f, 0.106714f, 0.037897f, 0.111461f,
        -0.024406f, 0.304769f, 0.254676f, -0.069235f,
        0.082566f, 0.147260f, 0.326969f, 0.148888f,
        0.055270f, -0.087985f, 0.261720f, 0.143442f,
        0.026812f, 0.238212f, 0.194020f, 0.056341f,
        -0.025854f, -0.034444f, -0.156631f, 0.205174f,
        0.089008f, -0.139811f, -0.100147f, -0.037830f,
        -0.029230f, -0.055641f, 0.033248f, -0.016512f,
        0.155244f, 0.247315f, -0.124694f, -0.048414f,
        -0.062219f, 0.193683f, 0.004574f, 0.055089f,
        0.093565f, 0.167712f, 0.167581f, 0.018895f,
        0.215258f, 0.122609f, 0.090520f, -0.067219f,
        -0.049029f, -0.099615f, 0.241804f, -0.094893f,
        -0.176248f, 0.001727f, -0.134473f, 0.104442f,
        0.050942f, 0.081165f, 0.072156f, 0.121646f,
        0.002656f, -0.297974f, -0.133587f, -0.060121f,
        -0.092515f, -0.048974f, -0.084754f, -0.180111f,
        -0.038590f, 0.086283f, -0.134636f, -0.107249f,
        0.132890f, 0.141556f, 0.249425f, 0.130273f,
        -0.030031f, 0.073212f, -0.008155f, 0.019931f,
        0.071688f, 0.000300f, -0.019525f, -0.021725f,
        -0.040993f, -0.086841f, 0.070124f, 0.240033f,
        0.265350f, 0.043208f, 0.166754f, 0.091453f,
        0.060916f, -0.036972f, -0.091043f, 0.079873f,
        0.219781f, 0.158102f, -0.140618f, -0.043016f,
        0.124802f, 0.093668f, 0.103208f, 0.094872f,
        0.080541f, 0.137711f, 0.160566f, -0.169231f,
        0.013983f, 0.309508f, -0.004217f, -0.057200f,
        -0.064489f, 0.014066f, 0.361009f, 0.251328f,
        -0.080983f, -0.044183f, 0.061436f, -0.037381f,
        -0.078786f, 0.030993f, 0.066314f, 0.037683f,
        0.152325f, -0.091683f, 0.070203f, 0.217856f,
        0.036435f, -0.076462f, 0.006254f, -0.094431f,
        0.154829f, -0.023038f, -0.196961f, -0.024594f,
        0.178465f, -0.050139f, -0.045932f, -0.000965f,
        0.109112f, 0.046165f, -0.159373f, -0.008713f,
        0.041307f, 0.097129f, -0.057211f, -0.064599f,
        0.077165f, 0.176167f, 0.138322f, 0.065753f,
        -0.104950f, 0.017933f, 0.136255f, -0.011598f,
        0.047007f, 0.080550f, 0.068619f, 0.084661f,
        -0.035493f, -0.091314f, -0.041411f, 0.060971f,
        -0.101912f, -0.079870f, -0.085977f, -0.022686f,
        0.079788f, -0.098064f, -0.054603f, 0.040383f,
        0.300794f, 0.128603f, 0.094844f, 0.047407f,
        0.101825f, 0.061832f, -0.162160f, -0.204553f,
        -0.035165f, 0.101450f, -0.016641f, -0.027140f,
        -0.134392f, -0.008743f, 0.102331f, 0.114853f,
        0.009644f, 0.062823f, 0.237339f, 0.167843f,
        0.053066f, -0.012592f, 0.043158f, 0.002305f,
        0.065001f, -0.038929f, -0.020356f, 0.152343f,
        0.043469f, -0.029967f, -0.042948f, 0.032481f,
        0.068488f, -0.110840f, -0.111083f, 0.111980f,
        -0.002072f, -0.005562f, 0.082926f, 0.006635f,
        -0.108153f, 0.024242f, -0.086464f, -0.189884f,
        -0.017492f, 0.191456f, -0.007683f, -0.128769f,
        -0.038017f, -0.132380f, 0.091926f, 0.079696f,
        -0.106728f, -0.007656f, 0.172744f, 0.011576f,
        0.009883f, 0.083258f, -0.026516f, 0.145534f,
        0.153924f, -0.130290f, -0.108945f, 0.124490f,
        -0.003186f, -0.100485f, 0.015024f, -0.060512f,
        0.026288f, -0.086713f, -0.169012f, 0.076517f,
        0.215778f, 0.043701f, -0.131642f, -0.012585f,
        -0.045181f, -0.118183f, -0.241544f, -0.167293f,
        -0.020107f, -0.019917f, -0.101827f, -0.107096f,
        -0.010503f, 0.044938f, 0.189680f, 0.217119f,
        -0.046086f, 0.044508f, 0.199716f, -0.036004f,
        -0.148927f, 0.013355f, -0.078279f, 0.030451f,
        0.056301f, -0.024609f, 0.083224f, 0.099533f,
        -0.039432f, -0.138880f, 0.005482f, -0.024120f,
        -0.140468f, -0.066381f, -0.017057f, 0.009260f,
        -0.058004f, -0.028486f, -0.061610f, 0.007483f,
        -0.158309f, -0.150687f, -0.044595f, -0.105121f,
        -0.045763f, -0.006618f, -0.024419f, -0.117713f,
        -0.119366f, -0.175941f, -0.071542f, 0.119027f,
        0.111362f, 0.043080f, 0.034889f, 0.093003f,
        0.007842f, 0.057368f, -0.108834f, -0.079968f,
        0.230959f, 0.020205f, 0.011470f, 0.098877f,
        0.101310f, -0.030215f, -0.018018f, -0.059552f,
        -0.106157f, 0.021866f, -0.036471f, 0.080051f,
        0.041165f, -0.082101f, 0.117726f, 0.030961f,
        -0.054763f, -0.084102f, -0.185778f, -0.061305f,
        -0.038089f, -0.110728f, -0.264010f, 0.076675f,
        -0.077111f, -0.137644f, 0.036232f, 0.277995f,
        0.019116f, 0.107738f, 0.144003f, 0.080304f,
        0.215036f, 0.228897f, 0.072713f, 0.077773f,
        0.120168f, 0.075324f, 0.062730f, 0.122478f,
        -0.049008f, 0.164912f, 0.162450f, 0.041246f,
        0.009891f, -0.097827f, -0.038700f, -0.023027f,
        -0.120020f, 0.203364f, 0.248474f, 0.149810f,
        -0.036276f, -0.082814f, -0.090343f, -0.027143f,
        -0.075689f, -0.320310f, -0.000500f, -0.143334f,
        -0.065077f, -0.186936f, 0.129372f, 0.116431f,
        0.181699f, 0.170436f, 0.418854f, 0.460045f,
        0.333719f, 0.230515f, 0.047822f, -0.044954f,
        -0.068086f, 0.140179f, -0.044821f, 0.085550f,
        0.092483f, -0.107296f, -0.130670f, -0.206629f,
        0.114601f, -0.317869f, -0.076663f, 0.038680f,
        0.212753f, -0.016059f, -0.126526f, -0.163602f,
        0.210154f, 0.099887f, -0.126366f, 0.118453f,
        0.019309f, -0.021611f, -0.096499f, -0.111809f,
        -0.200489f, 0.142854f, 0.228840f, -0.353346f,
        -0.179151f, 0.116834f, 0.252389f, -0.031728f,
        -0.188135f, -0.158998f, 0.386523f, 0.122315f,
        0.209944f, 0.394023f, 0.359030f, 0.260717f,
        0.170335f, 0.013683f, -0.142596f, -0.026138f,
        -0.011878f, -0.150519f, 0.047159f, -0.107062f,
        -0.147347f, -0.187689f, -0.186027f, -0.208048f,
        0.058468f, -0.073026f, -0.236556f, -0.079788f,
        -0.146216f, -0.058563f, -0.101361f, -0.071294f,
        -0.071093f, 0.116919f, 0.234304f, 0.306781f,
        0.321866f, 0.240000f, 0.073261f, -0.012173f,
        0.026479f, 0.050173f, 0.166127f, 0.228955f,
        0.061905f, 0.156460f, 0.205990f, 0.120672f,
        0.037350f, 0.167884f, 0.290099f, 0.420900f,
        -0.012601f, 0.189839f, 0.306378f, 0.118383f,
        -0.095598f, -0.072360f, -0.132496f, -0.224259f,
        -0.126021f, 0.022714f, 0.284039f, 0.051369f,
        -0.000927f, -0.058735f, -0.083354f, -0.141254f,
        -0.187578f, -0.202669f, 0.048902f, 0.246597f,
        0.441863f, 0.342519f, 0.066979f, 0.215286f,
        0.188191f, -0.072240f, -0.208142f, -0.030196f,
        0.178141f, 0.136985f, -0.043374f, -0.181098f,
        0.091815f, 0.116177f, -0.126690f, -0.386625f,
        0.368165f, 0.269149f, -0.088042f, -0.028823f,
        0.092961f, 0.024099f, 0.046112f, 0.176756f,
        0.135849f, 0.124955f, 0.195467f, -0.037218f,
        0.167217f, 0.188938f, 0.053528f, -0.066561f,
        0.133721f, -0.070565f, 0.115898f, 0.152435f,
        -0.116993f, -0.110592f, -0.179005f, 0.026668f,
        0.080530f, 0.075084f, -0.070401f, 0.012497f,
        0.021849f, -0.139764f, -0.022020f, -0.096301f,
        -0.064954f, -0.127446f, -0.013806f, -0.108315f,
        0.156285f, 0.149867f, -0.011382f, 0.064532f,
        0.029168f, 0.027393f, 0.069716f, 0.153735f,
        0.038459f, 0.230714f, 0.253840f, 0.059522f,
        -0.045053f, 0.014083f, 0.071103f, 0.068747f,
        0.095887f, 0.005832f, 0.144887f, 0.026357f,
        -0.067359f, -0.044151f, -0.123283f, -0.019911f,
        0.005318f, 0.109208f, -0.003201f, -0.021734f,
        0.142025f, -0.066907f, -0.120070f, -0.188639f,
        0.012472f, -0.048704f, -0.012366f, -0.184828f,
        0.168591f, 0.267166f, 0.058208f, -0.044101f,
        0.033500f, 0.178558f, 0.104550f, 0.122418f,
        0.080177f, 0.173246f, 0.298537f, 0.064173f,
        0.053397f, 0.174341f, 0.230984f, 0.117025f,
        0.166242f, 0.227781f, 0.120623f, 0.176952f,
        -0.011393f, -0.086483f, -0.008270f, 0.051700f,
        -0.153369f, -0.058837f, -0.057639f, -0.060115f,
        0.026349f, -0.160745f, -0.037894f, -0.048575f,
        0.041052f, -0.022112f, 0.060365f, 0.051906f,
        0.162657f, 0.138519f, -0.050185f, -0.005938f,
        0.071301f, 0.127686f, 0.062342f, 0.144400f,
        0.072600f, 0.198436f, 0.246219f, -0.078185f,
        -0.036169f, 0.075934f, 0.047328f, -0.013601f,
        0.087205f, 0.019900f, 0.022606f, -0.015365f,
        -0.092506f, 0.075275f, -0.116375f, 0.050500f,
        0.045118f, 0.166567f, 0.072073f, 0.060371f,
        0.131747f, -0.169863f, -0.039352f, -0.047486f,
        -0.039797f, -0.204312f, 0.021710f, 0.129443f,
        -0.021173f, 0.173416f, -0.070794f, -0.063986f,
        0.069689f, -0.064099f, -0.123201f, -0.017372f,
        -0.206870f, 0.065863f, 0.113226f, 0.024707f,
        -0.071341f, -0.066964f, -0.098278f, -0.062927f,
        0.075840f, 0.014716f, 0.019378f, 0.132699f,
        -0.074191f, -0.089557f, -0.078446f, -0.197488f,
        -0.173665f, 0.052583f, 0.044361f, 0.113549f,
        0.098492f, 0.077379f, -0.011146f, -0.192593f,
        -0.164435f, 0.045568f, 0.205699f, 0.049187f,
        -0.082281f, 0.134874f, 0.185499f, 0.034968f,
        -0.119561f, -0.112372f, -0.115091f, -0.054042f,
        -0.183816f, -0.078100f, 0.190695f, 0.091617f,
        0.004257f, -0.041135f, -0.061453f, -0.141592f,
        -0.194809f, -0.120638f, 0.020168f, 0.109672f,
        0.067398f, -0.015238f, -0.239145f, -0.264671f,
        -0.185176f, 0.050472f, 0.020793f, 0.035678f,
        0.022839f, -0.052055f, -0.127968f, -0.113049f,
        -0.228416f, -0.258281f, -0.053437f, 0.076424f,
        0.061450f, 0.237478f, 0.003618f, -0.055865f,
        -0.108087f, -0.028937f, 0.045585f, 0.052829f,
        -0.001471f, 0.022826f, 0.059565f, -0.104430f,
        -0.077266f, -0.211882f, -0.212078f, 0.028074f,
        0.075846f, 0.016265f, 0.161879f, 0.134477f,
        0.008935f, -0.048041f, 0.074692f, 0.004928f,
        -0.025156f, 0.192874f, 0.074410f, 0.308732f,
        0.267400f, 0.094208f, -0.005251f, 0.042041f,
        -0.032148f, 0.015588f, 0.252869f, 0.175302f,
        0.022892f, 0.081673f, 0.063208f, 0.162626f,
        0.194426f, 0.233890f, 0.262292f, 0.186930f,
        0.084079f, -0.286388f, -0.213034f, -0.048867f,
        -0.207669f, -0.170050f, 0.011673f, -0.092958f,
        -0.192786f, -0.273536f, 0.230904f, 0.266732f,
        0.320519f, 0.297155f, 0.548169f, 0.304922f,
        0.132687f, 0.247333f, 0.212488f, -0.271472f,
        -0.142105f, -0.002627f, -0.119215f, 0.128383f,
        0.100079f, -0.057490f, -0.121902f, -0.228892f,
        0.202292f, -0.399795f, -0.371326f, -0.095836f,
        -0.063626f, -0.161375f, -0.311180f, -0.294797f,
        0.242122f, 0.011788f, 0.095573f, 0.322523f,
        0.511840f, 0.322880f, 0.313259f, 0.173331f,
        0.002542f, -0.029802f, 0.324766f, -0.326170f,
        -0.340547f, -0.138288f, -0.002963f, -0.114060f,
        -0.377312f, -0.442570f, 0.212446f, -0.007759f,
        -0.011576f, 0.169711f, 0.308689f, 0.317348f,
        0.539390f, 0.332845f, 0.057331f, -0.068180f,
        0.101994f, 0.266995f, 0.209570f, 0.355730f,
        0.091635f, 0.170238f, 0.125215f, 0.274154f,
        0.070223f, 0.025515f, 0.049946f, -0.000550f,
        0.043715f, -0.141843f, 0.020844f, 0.129871f,
        0.256588f, 0.105015f, 0.148339f, 0.170682f,
        0.028792f, 0.074037f, 0.160042f, 0.405137f,
        0.246187f, 0.352160f, 0.168951f, 0.222263f,
        0.264439f, 0.065945f, 0.021963f, -0.075084f,
        0.093105f, 0.027318f, 0.098864f, 0.057566f,
        -0.080282f, 0.185032f, 0.314419f, 0.333727f,
        0.125798f, 0.294919f, 0.386002f, 0.217619f,
        -0.183517f, -0.278622f, -0.002342f, -0.027821f,
        -0.134266f, -0.331843f, -0.008296f, 0.124564f,
        0.053712f, -0.369016f, -0.095036f, 0.209381f,
        0.423760f, 0.371760f, 0.106397f, 0.369408f,
        0.485608f, 0.231201f, -0.138685f, -0.349208f,
        -0.070083f, 0.028991f, -0.081630f, -0.395992f,
        -0.146791f, -0.027354f, 0.063396f, -0.272484f,
        0.058299f, 0.338207f, 0.110767f, -0.052642f,
        -0.233848f, -0.027448f, 0.030328f, 0.155572f,
        -0.093826f, 0.019331f, 0.120638f, 0.006292f,
        -0.106083f, -0.236290f, -0.140933f, -0.088067f,
        -0.025138f, -0.208395f, -0.025502f, 0.144192f,
        -0.048353f, -0.106144f, -0.305121f, -0.114147f,
        0.090963f, 0.327727f, 0.035606f, -0.093779f,
        0.002651f, -0.171081f, -0.188131f, -0.216571f,
        -0.209101f, -0.054402f, 0.157147f, -0.057127f,
        0.066584f, 0.008988f, 0.041191f, 0.034456f,
        -0.078255f, 0.052099f, -0.022239f, 0.066981f,
        -0.117520f, -0.072637f, 0.062512f, 0.037570f,
        -0.057544f, -0.312359f, 0.034357f, -0.031549f,
        0.002566f, -0.207375f, -0.070654f, -0.018786f,
        -0.044815f, -0.012814f, -0.076320f, 0.078183f,
        0.023877f, 0.117078f, 0.022292f, -0.205424f,
        -0.060430f, -0.017296f, -0.004827f, -0.321036f,
        -0.092155f, 0.038837f, 0.073190f, -0.067513f,
        0.026521f, 0.171945f, 0.087318f, 0.034495f,
        -0.034089f, 0.154410f, -0.061431f, 0.007435f,
        -0.111094f, -0.095976f, 0.014741f, -0.132324f,
        -0.029517f, -0.192160f, 0.098667f, 0.020762f,
        0.177050f, -0.064510f, -0.054437f, -0.058678f,
        -0.001858f, 0.167602f, 0.015735f, 0.054338f,
        0.016477f, 0.186381f, -0.010667f, 0.054692f,
        0.126742f, 0.013140f, 0.090353f, -0.133608f,
        -0.018017f, -0.152619f, 0.027600f, -0.138700f,
        -0.050274f, 0.045141f, -0.118731f, 0.094797f,
        -0.167605f, 0.097461f, -0.009131f, 0.199920f,
        -0.052976f, 0.158194f, 0.178568f, -0.107600f,
        0.009671f, -0.084072f, -0.040258f, -0.205673f,
        0.102891f, 0.223511f, 0.042699f, 0.118548f,
        -0.021274f, 0.110997f, -0.155121f, 0.027696f,
        -0.149968f, 0.051552f, -0.129219f, 0.173524f,
        0.073972f, -0.189045f, -0.034523f, -0.106655f,
        -0.011843f, -0.197381f, 0.219413f, 0.183197f,
        -0.054920f, 0.144955f, 0.036517f, -0.085412f,
        -0.229070f, -0.143710f, -0.049486f, 0.156634f,
        -0.008673f, -0.064778f, 0.082344f, 0.145673f,
        0.002912f, -0.210121f, -0.116564f, 0.078425f,
        0.220908f, -0.067594f, 0.048610f, 0.084912f,
        -0.066202f, -0.112515f, -0.217767f, -0.082640f,
        -0.017414f, 0.230265f, -0.070735f, 0.066073f,
        0.215256f, 0.071157f, -0.087220f, -0.202235f,
        -0.011918f, 0.099562f, 0.174716f, -0.063845f,
        -0.121055f, 0.014367f, 0.132709f, -0.005060f,
        -0.244606f, -0.179693f, -0.134690f, 0.023239f,
        -0.193116f, -0.076975f, -0.021164f, -0.001938f,
        -0.163799f, -0.111437f, -0.210362f, -0.166376f,
        0.034754f, 0.010036f, -0.021917f, 0.068014f,
        -0.086893f, -0.251746f, -0.267171f, 0.037383f,
        0.003966f, 0.033571f, -0.151506f, 0.025437f,
        -0.020626f, -0.308454f, -0.343143f, -0.092263f,
        -0.026261f, -0.028345f, 0.036036f, 0.035169f,
        0.129470f, 0.122205f, 0.015661f, -0.070612f,
        -0.094333f, -0.066055f, -0.041083f, 0.159146f,
        0.073184f, 0.110044f, 0.174471f, 0.078069f,
        -0.014881f, 0.008116f, 0.013209f, 0.075857f,
        0.195605f, 0.062714f, 0.067955f, 0.056544f,
        -0.153908f, -0.141749f, -0.072550f, 0.033523f,
        -0.024665f, 0.134487f, 0.079076f, 0.133562f,
        0.227130f, 0.018054f, 0.004928f, 0.169162f,
        0.065152f, 0.072160f, 0.131631f, 0.096303f,
        0.054288f, 0.106256f, 0.114632f, 0.119038f,
        0.515200f, 0.247429f, 0.199134f, 0.211957f,
        0.127558f, -0.294684f, -0.194890f, -0.049988f,
        -0.112247f, -0.008122f, -0.006176f, 0.037035f,
        -0.110881f, -0.249989f, 0.152434f, 0.234621f,
        0.153340f, 0.349283f, 0.683049f, 0.157174f,
        0.124844f, 0.099136f, 0.064407f, -0.248400f,
        -0.155323f, -0.026498f, -0.023450f, 0.049051f,
        -0.114187f, 0.007195f, -0.176825f, -0.376926f,
        0.366159f, -0.179938f, -0.148508f, 0.006043f,
        0.170048f, 0.097866f, -0.102658f, -0.260430f,
        0.248868f, 0.037019f, -0.118111f, 0.078176f,
        0.194171f, 0.211328f, 0.368612f, 0.361213f,
        0.130013f, 0.094650f, 0.227396f, -0.178058f,
        -0.114782f, -0.008093f, 0.231080f, -0.011843f,
        -0.097917f, -0.325788f, 0.141879f, 0.119738f,
        -0.230427f, -0.117419f, -0.114153f, 0.037903f,
        0.116383f, 0.218773f, -0.101884f, 0.059466f,
        0.119255f, 0.010874f, -0.031449f, 0.045996f,
        0.119931f, 0.273760f, 0.311700f, 0.261794f,
        0.194809f, 0.339829f, 0.239449f, 0.064140f,
        0.077597f, 0.098996f, 0.143534f, 0.184602f,
        0.037507f, 0.225494f, 0.096142f, -0.147370f,
        -0.207833f, -0.174742f, -0.086391f, -0.038942f,
        0.159577f, -0.088492f, -0.000989f, 0.108154f,
        -0.025890f, -0.072713f, 0.025997f, -0.006803f,
        -0.086879f, -0.011290f, -0.269200f, -0.103450f,
        -0.124910f, -0.116340f, 0.141459f, 0.208800f,
        0.042268f, 0.265034f, 0.516474f, 0.217591f,
        -0.018843f, -0.313328f, -0.168363f, 0.047129f,
        0.090480f, -0.109852f, -0.018761f, 0.210669f,
        0.281269f, -0.043591f, -0.034147f, -0.237772f,
        -0.134843f, -0.072481f, -0.103831f, 0.038355f,
        0.308619f, 0.148023f, -0.045867f, -0.123950f,
        -0.210860f, -0.064973f, -0.036308f, -0.046731f,
        -0.022099f, 0.095776f, 0.409423f, 0.060635f,
        -0.065196f, 0.051828f, 0.027981f, -0.009609f,
        -0.137681f, -0.095011f, -0.019045f, 0.177278f,
        0.009759f, -0.092119f, -0.016958f, -0.133860f,
        -0.118421f, -0.032039f, -0.006214f, -0.084541f,
        0.063971f, -0.073642f, 0.165676f, 0.110443f,
        0.044131f, 0.046568f, 0.053292f, -0.055466f,
        0.015512f, 0.371947f, 0.232102f, -0.016923f,
        0.103979f, -0.091758f, 0.005907f, 0.209100f,
        0.157433f, 0.030518f, 0.250366f, 0.062322f,
        0.036720f, 0.094676f, 0.017306f, -0.010328f,
        -0.079012f, 0.016781f, -0.112435f, 0.061795f,
        0.042543f, -0.126799f, -0.009975f, -0.056760f,
        0.046424f, -0.194712f, -0.139399f, -0.037731f,
        0.157989f, -0.016261f, 0.123345f, 0.230563f,
        0.083300f, -0.016392f, 0.059567f, -0.016035f,
        -0.064767f, 0.231945f, 0.156629f, 0.034602f,
        0.145628f, 0.041315f, 0.034535f, 0.019967f,
        -0.089188f, -0.012091f, 0.307857f, 0.211405f,
        -0.025091f, -0.148249f, -0.129384f, 0.063536f,
        -0.068603f, -0.067941f, -0.035104f, 0.210832f,
        0.063810f, 0.062764f, -0.089889f, -0.030554f,
        0.014791f, -0.053362f, -0.037818f, -0.196640f,
        0.008388f, -0.082654f, 0.143056f, 0.064221f,
        0.069795f, 0.191040f, 0.097321f, -0.028679f,
        0.075794f, 0.313154f, 0.086240f, 0.207643f,
        0.017809f, 0.122867f, 0.224586f, 0.167403f,
        -0.023884f, 0.047434f, 0.344091f, 0.187745f,
        0.136177f, 0.141738f, 0.063799f, 0.045233f,
        -0.077342f, -0.003525f, -0.165041f, -0.025616f,
        -0.073745f, 0.164439f, 0.011200f, -0.145896f,
        -0.027954f, -0.061987f, -0.039874f, -0.142775f,
        0.151042f, -0.038238f, 0.053152f, 0.078615f,
        0.086061f, 0.100593f, 0.128046f, -0.071006f,
        -0.116558f, 0.208445f, 0.051086f, 0.076843f,
        0.023191f, -0.084781f, -0.011790f, 0.147807f,
        -0.048554f, -0.113932f, 0.283322f, 0.190934f,
        0.092789f, 0.033018f, -0.142428f, -0.142480f,
        -0.099023f, -0.041020f, -0.042760f, 0.203295f,
        -0.053475f, 0.042424f, 0.222839f, -0.019167f,
        -0.133176f, -0.276216f, -0.031998f, 0.117290f,
        0.177827f, -0.059973f, -0.064744f, -0.117040f,
        -0.155482f, -0.099531f, 0.164121f, -0.026682f,
        -0.093810f, 0.238993f, -0.006506f, 0.007830f,
        0.065819f, -0.203643f, -0.100925f, -0.053652f,
        -0.130770f, 0.026277f, 0.131796f, 0.032742f,
        0.127186f, 0.116694f, -0.161122f, -0.279773f,
        -0.252515f, -0.002638f, 0.042812f, 0.096776f,
        -0.123280f, 0.064858f, -0.010455f, -0.219760f,
        -0.239331f, -0.104363f, -0.058022f, -0.053584f,
        0.025611f, 0.005129f, -0.100418f, -0.045712f,
        -0.194418f, -0.126366f, -0.030530f, 0.051168f,
        0.215959f, 0.172402f, -0.054700f, -0.185995f,
        -0.278360f, -0.193693f, -0.040309f, 0.003735f,
        -0.007770f, 0.123556f, 0.190179f, -0.077315f,
        0.117403f, 0.212942f, 0.012160f, 0.000113f,
        0.027331f, 0.040202f, 0.033293f, 0.219438f,
        0.184174f, 0.259349f, 0.311206f, 0.082547f,
        -0.047875f, -0.078417f, 0.010746f, 0.082620f,
        0.311931f, 0.307605f, 0.003863f, 0.021405f,
        -0.026388f, -0.019572f, 0.020582f, -0.059353f,
        0.025199f, 0.261319f, 0.086316f, 0.143614f,
        0.107780f, 0.003900f, -0.188397f, -0.038563f,
        -0.106045f, -0.125154f, -0.010509f, 0.054021f,
        0.242130f, 0.279152f, 0.215546f, 0.346995f,
        0.440856f, 0.237452f, 0.234154f, 0.301646f,
        0.168929f, -0.208358f, -0.126848f, 0.010260f,
        0.121018f, -0.062975f, -0.052848f, 0.050341f,
        -0.061103f, -0.266482f, 0.107186f, 0.140221f,
        0.280065f, 0.287889f, 0.373198f, 0.151596f,
        0.013593f, 0.115616f, 0.014616f, -0.281710f,
        -0.237597f, -0.117305f, -0.000034f, -0.136739f,
        -0.196275f, -0.095225f, -0.125310f, -0.250514f,
        0.236804f, -0.071805f, -0.037421f, 0.048230f,
        0.321596f, 0.063632f, 0.024039f, -0.029133f,
        0.230983f, 0.160593f, -0.154355f, -0.013086f,
        -0.079929f, 0.094692f, 0.160391f, 0.180239f,
        0.053895f, 0.100759f, 0.288631f, 0.038191f,
        0.181692f, 0.229682f, 0.440166f, 0.063401f,
        0.006273f, 0.020865f, 0.338695f, 0.256244f,
        -0.043927f, 0.115617f, 0.003296f, 0.173965f,
        0.021318f, -0.040936f, -0.118932f, 0.182380f,
        0.235922f, -0.053233f, -0.015053f, -0.101057f,
        0.095341f, 0.051111f, 0.161831f, 0.032614f,
        0.159496f, 0.072375f, 0.025089f, 0.023748f,
        0.029151f, 0.161284f, -0.117717f, -0.036191f,
        -0.176822f, -0.162006f, 0.226542f, -0.078329f,
        0.043079f, -0.119172f, 0.054614f, -0.101365f,
        -0.064541f, -0.115304f, 0.135170f, 0.298872f,
        0.098060f, 0.089428f, -0.007497f, 0.110391f,
        -0.028824f, 0.020835f, -0.036804f, 0.125411f,
        0.192105f, -0.048931f, 0.003086f, -0.010681f,
        0.074698f, -0.016263f, 0.096063f, 0.060267f,
        -0.007277f, 0.139139f, -0.080635f, 0.036628f,
        0.086058f, 0.131979f, 0.085707f, 0.025301f,
        0.226094f, 0.194759f, 0.042193f, -0.157846f,
        -0.068402f, -0.141450f, -0.112659f, -0.076305f,
        -0.069085f, -0.114332f, -0.102005f, 0.132193f,
        -0.067042f, 0.106643f, 0.198964f, 0.171616f,
        0.167237f, -0.033730f, -0.026755f, 0.083621f,
        0.149459f, -0.002799f, -0.000318f, 0.011753f,
        0.065889f, -0.089375f, -0.049610f, 0.224579f,
        0.216548f, -0.034908f, -0.017851f, -0.088144f,
        0.007530f, 0.240268f, 0.073270f, 0.013263f,
        0.175323f, 0.012082f, 0.093993f, 0.015282f,
        0.105854f, 0.107990f, 0.077798f, -0.096166f,
        -0.079607f, 0.177820f, 0.142392f, 0.033337f,
        -0.078100f, -0.081616f, -0.046993f, 0.139459f,
        0.020272f, -0.123161f, 0.175269f, 0.105217f,
        0.057328f, 0.080909f, -0.012612f, -0.097081f,
        0.082060f, -0.096716f, -0.063921f, 0.201884f,
        0.128166f, -0.035051f, -0.032227f, -0.068139f,
        -0.115915f, 0.095080f, -0.086007f, -0.067543f,
        0.030776f, 0.032712f, 0.088937f, 0.054336f,
        -0.039329f, -0.114022f, 0.171672f, -0.112321f,
        -0.217646f, 0.065186f, 0.060223f, 0.192174f,
        0.055580f, -0.131107f, -0.144338f, 0.056730f,
        -0.034707f, -0.081616f, -0.135298f, -0.000614f,
        0.087189f, 0.014614f, 0.067709f, 0.107689f,
        0.225780f, 0.084361f, -0.008544f, 0.051649f,
        -0.048369f, -0.037739f, -0.060710f, 0.002654f,
        0.016935f, 0.085563f, -0.015961f, -0.019265f,
        0.111788f, 0.062376f, 0.202019f, 0.047713f,
        0.042261f, 0.069716f, 0.242913f, 0.021052f,
        -0.072812f, -0.155920f, -0.026436f, 0.035621f,
        -0.079300f, -0.028787f, -0.048329f, 0.084718f,
        -0.060565f, -0.083750f, -0.164075f, -0.040742f,
        -0.086219f, 0.015271f, -0.005204f, -0.016038f,
        0.045816f, -0.050433f, -0.077652f, 0.117109f,
        0.009611f, -0.009045f, -0.008634f, -0.055373f,
        -0.085968f, 0.028527f, -0.054736f, -0.168089f,
        0.175839f, 0.071205f, -0.023603f, 0.037907f,
        -0.004561f, -0.022634f, 0.123831f, 0.094469f,
        -0.072920f, -0.133642f, -0.014032f, -0.142754f,
        -0.026999f, -0.199409f, 0.013268f, 0.226989f,
        0.048650f, -0.170988f, -0.050141f, 0.007880f,
        0.061880f, 0.019078f, -0.043578f, -0.038139f,
        0.134814f, 0.054097f, -0.081670f, 0.176838f,
        0.047920f, -0.038176f, 0.050406f, -0.107181f,
        -0.036279f, 0.027060f, 0.081594f, -0.002820f,
        0.090507f, -0.033338f, -0.059571f, 0.013404f,
        -0.099860f, 0.073371f, 0.342805f, 0.098305f,
        -0.150910f, -0.020822f, -0.056960f, 0.046262f,
        -0.043413f, -0.149405f, -0.129105f, -0.010899f,
        -0.014229f, -0.179949f, -0.113044f, -0.049468f,
        -0.065513f, 0.090269f, -0.011919f, 0.087846f,
        0.095796f, 0.146127f, 0.101599f, 0.078066f,
        -0.084348f, -0.100002f, -0.020134f, -0.050169f,
        0.062122f, 0.014640f, 0.019143f, 0.036543f,
        0.180924f, -0.013976f, -0.066768f, -0.001090f,
        -0.070419f, -0.004839f, -0.001504f, 0.034483f,
        -0.044954f, -0.050336f, -0.088638f, -0.174782f,
        -0.116082f, -0.205507f, 0.015587f, -0.042839f,
        -0.096879f, -0.144097f, -0.050268f, -0.196796f,
        0.109639f, 0.271411f, 0.173732f, 0.108070f,
        0.156437f, 0.124255f, 0.097242f, 0.238693f,
        0.083941f, 0.109105f, 0.223940f, 0.267188f,
        0.027385f, 0.025819f, 0.125070f, 0.093738f,
        0.040353f, 0.038645f, -0.012730f, 0.144063f,
        0.052931f, -0.009138f, 0.084193f, 0.160272f,
        -0.041366f, 0.011951f, -0.121446f, -0.106713f,
        -0.047566f, 0.047984f, -0.255224f, -0.076116f,
        0.098685f, -0.150845f, -0.171513f, -0.156590f,
        0.058331f, 0.187493f, 0.413018f, 0.554265f,
        0.372242f, 0.237943f, 0.124571f, 0.110829f,
        0.010322f, -0.174477f, -0.067627f, -0.001979f,
        0.142913f, 0.040597f, 0.019907f, 0.025963f,
        -0.043585f, -0.120732f, 0.099937f, 0.091059f,
        0.247307f, 0.204226f, -0.042753f, -0.068580f,
        -0.119002f, 0.026722f, 0.034853f, -0.060934f,
        -0.025054f, -0.093026f, -0.035372f, -0.233209f,
        -0.049869f, -0.039151f, -0.022279f, -0.065380f,
        -9.063785f};
    return std::vector<float>(detector, detector + sizeof(detector)/sizeof(detector[0]));
}

class HOGConfInvoker :
    public ParallelLoopBody
{
public:
    HOGConfInvoker( const HOGDescriptor* _hog, const Mat& _img,
        double _hitThreshold, const Size& _padding,
        std::vector<DetectionROI>* locs,
        std::vector<Rect>* _vec, Mutex* _mtx )
    {
        hog = _hog;
        img = _img;
        hitThreshold = _hitThreshold;
        padding = _padding;
        locations = locs;
        vec = _vec;
        mtx = _mtx;
    }

    void operator()(const Range& range) const CV_OVERRIDE
    {
        CV_INSTRUMENT_REGION();

        int i, i1 = range.start, i2 = range.end;

        Size maxSz(cvCeil(img.cols/(*locations)[0].scale), cvCeil(img.rows/(*locations)[0].scale));
        Mat smallerImgBuf(maxSz, img.type());
        std::vector<Point> dets;

        for( i = i1; i < i2; i++ )
        {
            double scale = (*locations)[i].scale;

            Size sz(cvRound(img.cols / scale), cvRound(img.rows / scale));
            Mat smallerImg(sz, img.type(), smallerImgBuf.ptr());

            if( sz == img.size() )
                smallerImg = Mat(sz, img.type(), img.data, img.step);
            else
                resize(img, smallerImg, sz, 0, 0, INTER_LINEAR_EXACT);

            hog->detectROI(smallerImg, (*locations)[i].locations, dets, (*locations)[i].confidences, hitThreshold, Size(), padding);
            Size scaledWinSize = Size(cvRound(hog->winSize.width*scale), cvRound(hog->winSize.height*scale));
            mtx->lock();
            for( size_t j = 0; j < dets.size(); j++ )
                vec->push_back(Rect(cvRound(dets[j].x*scale),
                                    cvRound(dets[j].y*scale),
                                    scaledWinSize.width, scaledWinSize.height));
            mtx->unlock();
        }
    }

    const HOGDescriptor* hog;
    Mat img;
    double hitThreshold;
    std::vector<DetectionROI>* locations;
    Size padding;
    std::vector<Rect>* vec;
    Mutex* mtx;
};

void HOGDescriptor::detectROI(InputArray _img, const std::vector<cv::Point> &locations,
    CV_OUT std::vector<cv::Point>& foundLocations, CV_OUT std::vector<double>& confidences,
    double hitThreshold, cv::Size winStride, cv::Size padding) const
{
    CV_INSTRUMENT_REGION();

    Mat img = _img.getMat();
    foundLocations.clear();
    confidences.clear();

    if( svmDetector.empty() || locations.empty())
        return;

    if( winStride == Size() )
        winStride = cellSize;
    Size cacheStride(gcd(winStride.width, blockStride.width),
                     gcd(winStride.height, blockStride.height));

    size_t nwindows = locations.size();
    padding.width = (int)alignSize(std::max(padding.width, 0), cacheStride.width);
    padding.height = (int)alignSize(std::max(padding.height, 0), cacheStride.height);
    Size paddedImgSize(img.cols + padding.width*2, img.rows + padding.height*2);

    // HOGCache cache(this, img, padding, padding, nwindows == 0, cacheStride);
    HOGCache cache(this, img, padding, padding, true, cacheStride);
    if( !nwindows )
        nwindows = cache.windowsInImage(paddedImgSize, winStride).area();

    const HOGCache::BlockData* blockData = &cache.blockData[0];

    int nblocks = cache.nblocks.area();
    int blockHistogramSize = cache.blockHistogramSize;
    size_t dsize = getDescriptorSize();

    double rho = svmDetector.size() > dsize ? svmDetector[dsize] : 0;
    std::vector<float> blockHist(blockHistogramSize);

#if CV_SSE2 || CV_NEON
    float partSum[4];
#endif

    for( size_t i = 0; i < nwindows; i++ )
    {
        Point pt0;
        pt0 = locations[i];
        if( pt0.x < -padding.width || pt0.x > img.cols + padding.width - winSize.width ||
                pt0.y < -padding.height || pt0.y > img.rows + padding.height - winSize.height )
        {
            // out of image
            confidences.push_back(-10.0);
            continue;
        }

        double s = rho;
        const float* svmVec = &svmDetector[0];
        int j, k;

        for( j = 0; j < nblocks; j++, svmVec += blockHistogramSize )
        {
            const HOGCache::BlockData& bj = blockData[j];
            Point pt = pt0 + bj.imgOffset;

            // need to divide this into 4 parts!
            const float* vec = cache.getBlock(pt, &blockHist[0]);
#if CV_SSE2
            __m128 _vec = _mm_loadu_ps(vec);
            __m128 _svmVec = _mm_loadu_ps(svmVec);
            __m128 sum = _mm_mul_ps(_svmVec, _vec);

            for( k = 4; k <= blockHistogramSize - 4; k += 4 )
            {
                _vec = _mm_loadu_ps(vec + k);
                _svmVec = _mm_loadu_ps(svmVec + k);

                sum = _mm_add_ps(sum, _mm_mul_ps(_vec, _svmVec));
            }

            _mm_storeu_ps(partSum, sum);
            double t0 = partSum[0] + partSum[1];
            double t1 = partSum[2] + partSum[3];
            s += t0 + t1;
#elif CV_NEON
            float32x4_t _vec = vld1q_f32(vec);
            float32x4_t _svmVec = vld1q_f32(svmVec);
            float32x4_t sum = vmulq_f32(_svmVec, _vec);

            for( k = 4; k <= blockHistogramSize - 4; k += 4 )
            {
                _vec = vld1q_f32(vec + k);
                _svmVec = vld1q_f32(svmVec + k);

                sum = vaddq_f32(sum, vmulq_f32(_vec, _svmVec));
            }

            vst1q_f32(partSum, sum);
            double t0 = partSum[0] + partSum[1];
            double t1 = partSum[2] + partSum[3];
            s += t0 + t1;
#else
            for( k = 0; k <= blockHistogramSize - 4; k += 4 )
                s += vec[k]*svmVec[k] + vec[k+1]*svmVec[k+1] +
                        vec[k+2]*svmVec[k+2] + vec[k+3]*svmVec[k+3];
#endif
            for( ; k < blockHistogramSize; k++ )
                s += vec[k]*svmVec[k];
        }
        confidences.push_back(s);

        if( s >= hitThreshold )
            foundLocations.push_back(pt0);
    }
}

void HOGDescriptor::detectMultiScaleROI(InputArray _img,
    CV_OUT std::vector<cv::Rect>& foundLocations, std::vector<DetectionROI>& locations,
    double hitThreshold, int groupThreshold) const
{
    CV_INSTRUMENT_REGION();

    Mat img = _img.getMat();
    std::vector<Rect> allCandidates;
    Mutex mtx;

    parallel_for_(Range(0, (int)locations.size()),
                  HOGConfInvoker(this, img, hitThreshold, Size(8, 8),
                                 &locations, &allCandidates, &mtx));

    foundLocations.resize(allCandidates.size());
    std::copy(allCandidates.begin(), allCandidates.end(), foundLocations.begin());
    cv::groupRectangles(foundLocations, groupThreshold, 0.2);
}

void HOGDescriptor::groupRectangles(std::vector<cv::Rect>& rectList, std::vector<double>& weights, int groupThreshold, double eps) const
{
    CV_INSTRUMENT_REGION();

    if( groupThreshold <= 0 || rectList.empty() )
    {
        return;
    }

    CV_Assert(rectList.size() == weights.size());

    std::vector<int> labels;
    int nclasses = partition(rectList, labels, SimilarRects(eps));

    std::vector<cv::Rect_<double> > rrects(nclasses);
    std::vector<int> numInClass(nclasses, 0);
    std::vector<double> foundWeights(nclasses, -std::numeric_limits<double>::max());
    int i, j, nlabels = (int)labels.size();

    for( i = 0; i < nlabels; i++ )
    {
        int cls = labels[i];
        rrects[cls].x += rectList[i].x;
        rrects[cls].y += rectList[i].y;
        rrects[cls].width += rectList[i].width;
        rrects[cls].height += rectList[i].height;
        foundWeights[cls] = max(foundWeights[cls], weights[i]);
        numInClass[cls]++;
    }

    for( i = 0; i < nclasses; i++ )
    {
        // find the average of all ROI in the cluster
        cv::Rect_<double> r = rrects[i];
        double s = 1.0/numInClass[i];
        rrects[i] = cv::Rect_<double>(cv::saturate_cast<double>(r.x*s),
            cv::saturate_cast<double>(r.y*s),
            cv::saturate_cast<double>(r.width*s),
            cv::saturate_cast<double>(r.height*s));
    }

    rectList.clear();
    weights.clear();

    for( i = 0; i < nclasses; i++ )
    {
        cv::Rect r1 = rrects[i];
        int n1 = numInClass[i];
        double w1 = foundWeights[i];
        if( n1 <= groupThreshold )
            continue;
        // filter out small rectangles inside large rectangles
        for( j = 0; j < nclasses; j++ )
        {
            int n2 = numInClass[j];

            if( j == i || n2 <= groupThreshold )
                continue;

            cv::Rect r2 = rrects[j];

            int dx = cv::saturate_cast<int>( r2.width * eps );
            int dy = cv::saturate_cast<int>( r2.height * eps );

            if( r1.x >= r2.x - dx &&
                r1.y >= r2.y - dy &&
                r1.x + r1.width <= r2.x + r2.width + dx &&
                r1.y + r1.height <= r2.y + r2.height + dy &&
                (n2 > std::max(3, n1) || n1 < 3) )
                break;
        }

        if( j == nclasses )
        {
            rectList.push_back(r1);
            weights.push_back(w1);
        }
    }
}
}