roi_align_op.cu 13.2 KB
Newer Older
J
jerrywgz 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
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.

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. */

#include "paddle/fluid/operators/roi_align_op.h"
#include "paddle/fluid/platform/cuda_primitives.h"

namespace paddle {
namespace operators {

using Tensor = framework::Tensor;
using LoDTensor = framework::LoDTensor;

static constexpr int kNumCUDAThreads = 512;
static constexpr int kNumMaxinumNumBlocks = 4096;

static inline int NumBlocks(const int N) {
  return std::min((N + kNumCUDAThreads - 1) / kNumCUDAThreads,
                  kNumMaxinumNumBlocks);
}

#define CUDA_1D_KERNEL_LOOP(i, n)                              \
  for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < (n); \
       i += blockDim.x * gridDim.x)

template <class T>
J
jerrywgz 已提交
37 38
__device__ T BilinearInterpolate(const T* input_data, const int height,
                                 const int width, T y, T x) {
J
jerrywgz 已提交
39 40 41
  if (y < -1.0 || y > height || x < -1.0 || x > width) {
    return 0;
  }
J
jerrywgz 已提交
42 43
  y = y <= 0 ? 0 : y;
  x = x <= 0 ? 0 : x;
J
jerrywgz 已提交
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
  int y_low = static_cast<int>(y);
  int x_low = static_cast<int>(x);
  int y_high;
  int x_high;
  if (y_low >= height - 1) {
    y_high = y_low = height - 1;
    y = static_cast<T>(y_low);
  } else {
    y_high = y_low + 1;
  }
  if (x_low >= width - 1) {
    x_high = x_low = width - 1;
    x = static_cast<T>(x_low);
  } else {
    x_high = x_low + 1;
  }
  T ly = y - y_low, lx = x - x_low;
  T hy = 1. - ly, hx = 1. - lx;

  T v1 = input_data[y_low * width + x_low];
  T v2 = input_data[y_low * width + x_high];
  T v3 = input_data[y_high * width + x_low];
  T v4 = input_data[y_high * width + x_high];
  T w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx;

  T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4);
  return val;
}

template <class T>
J
jerrywgz 已提交
74 75 76 77
__device__ void BilinearInterpolateGradient(const int height, const int width,
                                            T y, T x, T* w1, T* w2, T* w3,
                                            T* w4, int* x_low, int* x_high,
                                            int* y_low, int* y_high) {
J
jerrywgz 已提交
78 79 80 81
  if (y < -1.0 || y > height || x < -1.0 || x > width) {
    return;
  }

J
jerrywgz 已提交
82 83
  y = y <= 0 ? 0 : y;
  x = x <= 0 ? 0 : x;
84 85 86 87 88
  *y_low = static_cast<int>(y);
  *x_low = static_cast<int>(x);
  if (*y_low >= height - 1) {
    *y_high = *y_low = height - 1;
    y = static_cast<T>(*y_low);
J
jerrywgz 已提交
89
  } else {
90
    *y_high = *y_low + 1;
J
jerrywgz 已提交
91
  }
92 93 94
  if (*x_low >= width - 1) {
    *x_high = *x_low = width - 1;
    x = static_cast<T>(*x_low);
J
jerrywgz 已提交
95
  } else {
96
    *x_high = *x_low + 1;
J
jerrywgz 已提交
97
  }
98
  T ly = y - *y_low, lx = x - *x_low;
J
jerrywgz 已提交
99
  T hy = 1. - ly, hx = 1. - lx;
100
  *w1 = hy * hx, *w2 = hy * lx, *w3 = ly * hx, *w4 = ly * lx;
J
jerrywgz 已提交
101 102 103 104 105 106 107 108 109

  return;
}

template <class T>
__global__ void GPUROIAlignForward(
    const int nthreads, const T* input_data, const T* input_rois,
    const float spatial_scale, const int channels, const int height,
    const int width, const int pooled_height, const int pooled_width,
110
    const int sampling_ratio, int* roi_batch_id_data, T* output_data) {
J
jerrywgz 已提交
111 112 113 114 115 116 117 118 119 120 121 122 123 124
  CUDA_1D_KERNEL_LOOP(i, nthreads) {
    int pw = i % pooled_width;
    int ph = (i / pooled_width) % pooled_height;
    int c = (i / pooled_width / pooled_height) % channels;
    int n = i / pooled_width / pooled_height / channels;

    const T* offset_input_rois = input_rois + n * kROISize;
    int roi_batch_ind = roi_batch_id_data[n];

    T roi_xmin = offset_input_rois[0] * spatial_scale;
    T roi_ymin = offset_input_rois[1] * spatial_scale;
    T roi_xmax = offset_input_rois[2] * spatial_scale;
    T roi_ymax = offset_input_rois[3] * spatial_scale;

125 126
    T roi_width = max(roi_xmax - roi_xmin, static_cast<T>(1.));
    T roi_height = max(roi_ymax - roi_ymin, static_cast<T>(1.));
J
jerrywgz 已提交
127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
    T bin_size_h = static_cast<T>(roi_height) / static_cast<T>(pooled_height);
    T bin_size_w = static_cast<T>(roi_width) / static_cast<T>(pooled_width);

    const T* offset_input_data =
        input_data + (roi_batch_ind * channels + c) * height * width;

    int roi_bin_grid_h = (sampling_ratio > 0)
                             ? sampling_ratio
                             : ceil(roi_height / pooled_height);
    int roi_bin_grid_w =
        (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);
    const T count = roi_bin_grid_h * roi_bin_grid_w;
    T output_val = 0;
    for (int iy = 0; iy < roi_bin_grid_h; iy++) {
      const T y = roi_ymin + ph * bin_size_h +
                  static_cast<T>(iy + .5f) * bin_size_h /
                      static_cast<T>(roi_bin_grid_h);
      for (int ix = 0; ix < roi_bin_grid_w; ix++) {
        const T x = roi_xmin + pw * bin_size_w +
                    static_cast<T>(ix + .5f) * bin_size_w /
                        static_cast<T>(roi_bin_grid_w);
J
jerrywgz 已提交
148
        T val = BilinearInterpolate(offset_input_data, height, width, y, x);
J
jerrywgz 已提交
149 150 151 152 153 154 155 156 157 158
        output_val += val;
      }
    }
    output_val /= count;
    output_data[i] = output_val;
  }
}

template <typename T>
__global__ void GPUROIAlignBackward(const int nthreads, const T* input_rois,
159
                                    const T* out_grad, const int num_rois,
J
jerrywgz 已提交
160 161 162 163 164 165 166 167 168
                                    const float spatial_scale,
                                    const int channels, const int height,
                                    const int width, const int pooled_height,
                                    const int pooled_width,
                                    const int sampling_ratio,
                                    int* roi_batch_id_data, T* input_grad) {
  CUDA_1D_KERNEL_LOOP(i, nthreads) {
    int pw = i % pooled_width;
    int ph = (i / pooled_width) % pooled_height;
169
    int c = (i / pooled_width / pooled_height) % channels;
J
jerrywgz 已提交
170 171 172 173 174 175 176 177 178
    int n = i / pooled_width / pooled_height / channels;
    const T* offset_input_rois = input_rois + n * kROISize;
    int roi_batch_ind = roi_batch_id_data[n];

    T roi_xmin = offset_input_rois[0] * spatial_scale;
    T roi_ymin = offset_input_rois[1] * spatial_scale;
    T roi_xmax = offset_input_rois[2] * spatial_scale;
    T roi_ymax = offset_input_rois[3] * spatial_scale;

179 180
    T roi_width = max(roi_xmax - roi_xmin, static_cast<T>(1.));
    T roi_height = max(roi_ymax - roi_ymin, static_cast<T>(1.));
J
jerrywgz 已提交
181 182 183
    T bin_size_h = static_cast<T>(roi_height) / static_cast<T>(pooled_height);
    T bin_size_w = static_cast<T>(roi_width) / static_cast<T>(pooled_width);

184
    T* offset_input_grad =
J
jerrywgz 已提交
185 186 187 188 189 190 191 192 193 194 195 196 197 198
        input_grad + (roi_batch_ind * channels + c) * height * width;

    const T* offset_out_grad =
        out_grad + (n * channels + c) * pooled_height * pooled_width;
    const T out_grad_this_bin = offset_out_grad[ph * pooled_width + pw];

    int roi_bin_grid_h = (sampling_ratio > 0)
                             ? sampling_ratio
                             : ceil(roi_height / pooled_height);
    int roi_bin_grid_w =
        (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);

    const T count = roi_bin_grid_h * roi_bin_grid_w;
    for (int iy = 0; iy < roi_bin_grid_h; iy++) {
199
      const T y = roi_ymin + ph * bin_size_h +
J
jerrywgz 已提交
200 201 202
                  static_cast<T>(iy + .5f) * bin_size_h /
                      static_cast<T>(roi_bin_grid_h);
      for (int ix = 0; ix < roi_bin_grid_w; ix++) {
203
        const T x = roi_xmin + pw * bin_size_w +
J
jerrywgz 已提交
204 205
                    static_cast<T>(ix + .5f) * bin_size_w /
                        static_cast<T>(roi_bin_grid_w);
206 207
        T w1 = 0, w2 = 0, w3 = 0, w4 = 0;
        int x_low = -1, x_high = -1, y_low = -1, y_high = -1;
J
jerrywgz 已提交
208 209
        BilinearInterpolateGradient(height, width, y, x, &w1, &w2, &w3, &w4,
                                    &x_low, &x_high, &y_low, &y_high);
J
jerrywgz 已提交
210 211 212 213 214 215 216 217 218 219 220 221
        T diff1 = out_grad_this_bin * w1 / count;
        T diff2 = out_grad_this_bin * w2 / count;
        T diff3 = out_grad_this_bin * w3 / count;
        T diff4 = out_grad_this_bin * w4 / count;
        if (x_low >= 0 && x_high >= 0 && y_low >= 0 && y_high >= 0) {
          platform::CudaAtomicAdd(offset_input_grad + y_low * width + x_low,
                                  diff1);
          platform::CudaAtomicAdd(offset_input_grad + y_low * width + x_high,
                                  diff2);
          platform::CudaAtomicAdd(offset_input_grad + y_high * width + x_low,
                                  diff3);
          platform::CudaAtomicAdd(offset_input_grad + y_high * width + x_high,
222
                                  diff4);
J
jerrywgz 已提交
223 224 225 226 227 228 229 230 231 232
        }
      }
    }
  }
}

template <typename Place, typename T>
class GPUROIAlignOpKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
233
    auto* in = ctx.Input<Tensor>("X");
J
jerrywgz 已提交
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
    auto* rois = ctx.Input<LoDTensor>("ROIs");
    auto* out = ctx.Output<Tensor>("Out");

    auto pooled_height = ctx.Attr<int>("pooled_height");
    auto pooled_width = ctx.Attr<int>("pooled_width");
    auto spatial_scale = ctx.Attr<float>("spatial_scale");
    auto sampling_ratio = ctx.Attr<int>("sampling_ratio");

    auto in_dims = in->dims();
    int batch_size = in_dims[0];
    int channels = in_dims[1];
    int height = in_dims[2];
    int width = in_dims[3];

    int rois_num = rois->dims()[0];

    if (rois_num == 0) return;

    int output_size = out->numel();
    int blocks = NumBlocks(output_size);
    int threads = kNumCUDAThreads;

    Tensor roi_batch_id_list;
    roi_batch_id_list.Resize({rois_num});
    int* roi_batch_id_data =
        roi_batch_id_list.mutable_data<int>(platform::CPUPlace());
    auto rois_lod = rois->lod().back();
    int rois_batch_size = rois_lod.size() - 1;
    PADDLE_ENFORCE_EQ(
        rois_batch_size, batch_size,
        "The rois_batch_size and imgs batch_size must be the same.");
    int rois_num_with_lod = rois_lod[rois_batch_size];
    PADDLE_ENFORCE_EQ(rois_num, rois_num_with_lod,
                      "The rois_num from input and lod must be the same.");
    for (int n = 0; n < rois_batch_size; ++n) {
      for (size_t i = rois_lod[n]; i < rois_lod[n + 1]; ++i) {
        roi_batch_id_data[i] = n;
      }
    }
    Tensor roi_batch_id_list_gpu;
J
jerrywgz 已提交
274 275
    framework::TensorCopySync(roi_batch_id_list, ctx.GetPlace(),
                              &roi_batch_id_list_gpu);
J
jerrywgz 已提交
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
    GPUROIAlignForward<
        T><<<blocks, threads, 0, ctx.cuda_device_context().stream()>>>(
        output_size, in->data<T>(), rois->data<T>(), spatial_scale, channels,
        height, width, pooled_height, pooled_width, sampling_ratio,
        roi_batch_id_list_gpu.data<int>(),
        out->mutable_data<T>(ctx.GetPlace()));
  }
};

template <typename Place, typename T>
class GPUROIAlignGradOpKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
    auto* in = ctx.Input<Tensor>("X");
    auto* rois = ctx.Input<LoDTensor>("ROIs");

    auto* out_grad = ctx.Input<Tensor>(framework::GradVarName("Out"));
    auto* in_grad = ctx.Output<Tensor>(framework::GradVarName("X"));

    auto pooled_height = ctx.Attr<int>("pooled_height");
    auto pooled_width = ctx.Attr<int>("pooled_width");
    auto spatial_scale = ctx.Attr<float>("spatial_scale");
    auto sampling_ratio = ctx.Attr<int>("sampling_ratio");

    int rois_num = rois->dims()[0];
    int channels = in->dims()[1];
    int height = in->dims()[2];
    int width = in->dims()[3];

J
jerrywgz 已提交
305 306 307 308 309 310 311 312 313 314 315 316
    if (!in_grad) {
      return;
    }
    Tensor roi_batch_id_list;
    roi_batch_id_list.Resize({rois_num});
    int* roi_batch_id_data =
        roi_batch_id_list.mutable_data<int>(platform::CPUPlace());
    auto rois_lod = rois->lod().back();
    int rois_batch_size = rois_lod.size() - 1;
    for (int n = 0; n < rois_batch_size; ++n) {
      for (size_t i = rois_lod[n]; i < rois_lod[n + 1]; ++i) {
        roi_batch_id_data[i] = n;
J
jerrywgz 已提交
317 318
      }
    }
J
jerrywgz 已提交
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
    Tensor roi_batch_id_list_gpu;
    framework::TensorCopySync(roi_batch_id_list, ctx.GetPlace(),
                              &roi_batch_id_list_gpu);

    in_grad->mutable_data<T>(ctx.GetPlace());
    math::SetConstant<Place, T> set_zero;
    set_zero(ctx.cuda_device_context(), in_grad, static_cast<T>(0));

    int output_grad_size = out_grad->numel();
    int blocks = NumBlocks(output_grad_size);
    int threads = kNumCUDAThreads;

    if (output_grad_size > 0) {
      GPUROIAlignBackward<
          T><<<blocks, threads, 0, ctx.cuda_device_context().stream()>>>(
          output_grad_size, rois->data<T>(), out_grad->data<T>(), rois_num,
          spatial_scale, channels, height, width, pooled_height, pooled_width,
          sampling_ratio, roi_batch_id_list_gpu.data<int>(),
          in_grad->mutable_data<T>(ctx.GetPlace()));
    }
J
jerrywgz 已提交
339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
  }
};

}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
REGISTER_OP_CUDA_KERNEL(
    roi_align,
    ops::GPUROIAlignOpKernel<paddle::platform::CUDADeviceContext, float>,
    ops::GPUROIAlignOpKernel<paddle::platform::CUDADeviceContext, double>);
REGISTER_OP_CUDA_KERNEL(
    roi_align_grad,
    ops::GPUROIAlignGradOpKernel<paddle::platform::CUDADeviceContext, float>,
    ops::GPUROIAlignGradOpKernel<paddle::platform::CUDADeviceContext, double>);