hl_cnn.h 9.3 KB
Newer Older
Z
zhangjinchao01 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
/* Copyright (c) 2016 Baidu, Inc. All Rights Reserve.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License. */


#ifndef HL_CNN_H_
#define HL_CNN_H_

#include "hl_base.h"

/**
 * @brief   Shrink column to feature.
 *
 * @param[in]   dataCol     expand data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   blockH      filter height.
 * @param[in]   blockW      filter width.
 * @param[in]   strideH     stride height.
 * @param[in]   strideW     stride width.
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
 * @param[in]   outputH     output height.
 * @param[in]   outputW     output width.
 * @param[out]  dataIm      output image data.
 * @param[in]   alpha
 * @param[in]   beta
 */
extern void hl_shrink_col2feature(
    const real * dataCol, size_t channels,
    size_t height, size_t width,
    size_t blockH, size_t blockW,
    size_t strideH, size_t strideW,
    size_t paddingH, size_t paddingW,
    size_t outputH, size_t outputW,
    real* dataIm,
    real alpha = 1.0f, real beta = 0.0f);

/**
 * @brief   Expand feature to column.
 *
 * @param[in]   dataIm      input image data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   blockH      filter height.
 * @param[in]   blockW      filter width.
 * @param[in]   strideH     stride height.
 * @param[in]   strideW     stride width.
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
 * @param[in]   outputH     output height.
 * @param[in]   outputW     output width.
 * @param[out]  dataCol     expand data.
 *
 */
extern void hl_expand_feature2col(
    const real* dataIm, size_t channels,
    size_t height, size_t width,
    size_t blockH, size_t blockW,
    size_t strideH, size_t strideW,
    size_t paddingH, size_t paddingW,
    size_t outputH, size_t outputW,
    real* dataCol);

/**
 * @brief   Maximum pool forward.
 *
 * @param[in]   frameCnt    batch size of input image.
 * @param[in]   inputData   input data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   pooledH     output image height.
 * @param[in]   pooledW     output image width.
87 88 89 90 91 92
 * @param[in]   sizeX       width of pooling window.
 * @param[in]   sizeY       height of pooling window.
 * @param[in]   strideH     pooling stride height.
 * @param[in]   strideW     pooling stride width.
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
Z
zhangjinchao01 已提交
93 94 95 96
 * @param[out]  tgtData     output data.
 *
 */
extern void hl_maxpool_forward(
97 98 99 100 101 102 103
    const int frameCnt, const real* inputData,
    const int channels,
    const int height, const int width,
    const int pooledH, const int pooledW,
    const int sizeX, const int sizeY,
    const int strideH, const int strideW,
    const int paddingH, const int paddingW, real* tgtData);
Z
zhangjinchao01 已提交
104 105 106 107 108 109 110 111 112 113 114 115 116

/**
 * @brief   Maximum pool backward.
 *
 * @param[in]   frameCnt    batch size of input image.
 * @param[in]   inputData   input data.
 * @param[out]  outData     output data.
 * @param[out]  outGrad     output grad data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   pooledH     output image height.
 * @param[in]   pooledW     output image width.
117 118 119 120
 * @param[in]   sizeX       width of pooling window.
 * @param[in]   sizeY       height of pooling window.
 * @param[in]   strideH     pooling stride height.
 * @param[in]   strideW     pooling stride width.
Z
zhangjinchao01 已提交
121 122
 * @param[in]   scaleA      scale.
 * @param[in]   scaleB      scale.
123 124 125
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
 * @param[out]  targetGrad  output grad.
Z
zhangjinchao01 已提交
126 127 128
 *
 */
extern void hl_maxpool_backward(
129
    const int frameCnt, const real* inputData,
Z
zhangjinchao01 已提交
130
    const real* outData, const real* outGrad,
131 132 133 134 135 136 137 138
    const int channels, const int height,
    const int width,
    const int pooledH, const int pooledW,
    const int sizeX, const int sizeY,
    const int strideH, const int strideW,
    const int paddingH, const int paddingW,
    real scaleA, real scaleB,
    real* targetGrad);
Z
zhangjinchao01 已提交
139 140 141 142 143 144 145 146 147 148 149

/**
 * @brief   Averge pool forward.
 *
 * @param[in]   frameCnt    batch size of input image.
 * @param[in]   inputData   input data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   pooledH     output image height.
 * @param[in]   pooledW     output image width.
150 151 152 153 154 155
 * @param[in]   sizeX       width of pooling window.
 * @param[in]   sizeY       height of pooling window.
 * @param[in]   strideH     pooling stride height.
 * @param[in]   strideW     pooling stride width.
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
Z
zhangjinchao01 已提交
156 157 158 159
 * @param[out]  tgtData     output data.
 *
 */
extern void hl_avgpool_forward(
160 161 162 163 164 165 166
    const int frameCnt, const real* inputData,
    const int channels,
    const int height, const int width,
    const int pooledH, const int pooledW,
    const int sizeX, const int sizeY,
    const int strideH, const int strideW,
    const int paddingH, const int paddingW, real* tgtData);
Z
zhangjinchao01 已提交
167 168 169 170 171

/**
 * @brief   Maximum pool backward.
 *
 * @param[in]   frameCnt    batch size of input image.
172
 * @param[in]   outGrad     output grad data.
Z
zhangjinchao01 已提交
173 174 175 176 177
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   pooledH     output image height.
 * @param[in]   pooledW     output image width.
178 179 180 181 182 183
 * @param[in]   sizeX       width of pooling window.
 * @param[in]   sizeY       height of pooling window.
 * @param[in]   strideH     pooling stride height.
 * @param[in]   strideW     pooling stride width.
 * @param[in]   paddingH    padding height.
 * @param[in]   paddingW    padding width.
Z
zhangjinchao01 已提交
184 185
 * @param[in]   scaleA      scale.
 * @param[in]   scaleB      scale.
186
 * @param[out]  backGrad    output grad.
Z
zhangjinchao01 已提交
187 188 189
 *
 */
extern void hl_avgpool_backward(
190 191 192 193 194 195 196 197 198
    const int frameCnt, const real* outGrad,
    const int channels, const int height,
    const int width,
    const int pooledH, const int pooledW,
    const int sizeX, const int sizeY,
    const int strideH, const int strideW,
    int paddingH, int paddingW,
    real scaleA, real scaleB,
    real* backGrad);
Z
zhangjinchao01 已提交
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242

/**
 * @brief   Cross-map-respose normalize forward.
 *
 * @param[in]   frameCnt    batch size of input image.
 * @param[in]   in          input data.
 * @param[in]   scale       buffer.
 * @param[out]  out         output data.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   sizeX       size.
 * @param[in]   alpha       scale.
 * @param[in]   beta        scale.
 *
 */
extern void hl_CMRNorm_forward(
    size_t frameCnt, const real* in, real* scale, real* out,
    size_t channels, size_t height, size_t width, size_t sizeX,
    real alpha, real beta);

/**
 * @brief   Cross-map-respose normalize backward.
 *
 * @param[in]   frameCnt    batch size of input image.
 * @param[in]   inV         input data.
 * @param[in]   scale       buffer.
 * @param[out]  outV        output value.
 * @param[out]  outDiff     output grad.
 * @param[out]  inDiff      input grad.
 * @param[in]   channels    number of channel.
 * @param[in]   height      image height.
 * @param[in]   width       image width.
 * @param[in]   sizeX       size.
 * @param[in]   alpha       scale.
 * @param[in]   beta        scale.
 *
 */
extern void hl_CMRNorm_backward(
    size_t frameCnt, const real* inV, const real* scale,
    const real* outV, const real* outDiff, real *inDiff,
    size_t channels, size_t height, size_t width, size_t sizeX,
    real alpha, real beta);

243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
/**
 * @brief   MaxOut forward.
 *
 * @param[in]   inData      input data.
 * @param[out]  outData     output data.
 * @param[out]  idData      output maxId.
 * @param[in]   batchSize   batchSize.
 * @param[in]   size        number of channels * image height * image width.
 * @param[in]   featLen     feature length = image height * image width.
 * @param[in]   groups      number of groups.
 */
extern void hl_maxout_forward(
    const real* inData, real* outData, int* idData,
    size_t batchSize, size_t size, size_t featLen, size_t groups);

/**
 * @brief   MaxOut backward.
 *
 * @param[out]  inGrad      input grad data.
 * @param[in]   outGrad     output grad data.
 * @param[in]   idData      output maxId.
 * @param[in]   batchSize   batchSize.
 * @param[in]   size        number of channels * image height * image width.
 * @param[in]   featLen     feature length = image height * image width.
 * @param[in]   groups      number of groups.
 */
extern void hl_maxout_backward(
    real* inGrad, const real* outGrad, const int* idData,
    size_t batchSize, size_t size, size_t featLen, size_t groups);

Z
zhangjinchao01 已提交
273
#endif /* HL_CNN_H_ */