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
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
|
#include "caffe2/core/context_gpu.h"
#include "caffe2/operators/resize_3d_op.h"
#include "caffe2/utils/GpuAtomics.cuh"
#include "caffe2/utils/math.h"
namespace caffe2 {
namespace {
__global__ void NearestNeighbor3DKernel(
const int size,
const int num_channels,
const int input_frames,
const int input_height,
const int input_width,
const int output_frames,
const int output_height,
const int output_width,
const float temporal_scale,
const float height_scale,
const float width_scale,
const float* X,
float* Y) {
CUDA_1D_KERNEL_LOOP(index, size) {
int indexTemp = index;
const int w = indexTemp % output_width;
indexTemp /= output_width;
const int h = indexTemp % output_height;
indexTemp /= output_height;
const int f = indexTemp % output_frames;
indexTemp /= output_frames;
const int c = indexTemp % num_channels;
indexTemp /= num_channels;
const int n = indexTemp;
const int in_f = fminf(f / temporal_scale, input_frames - 1);
const int in_y = fminf(h / height_scale, input_height - 1);
const int in_x = fminf(w / width_scale, input_width - 1);
Y[index] =
X[(((n * num_channels + c) * input_frames + in_f) * input_height + in_y)
* input_width + in_x];
}
}
__global__ void NearestNeighbor3DGradientKernel(
const int size,
const int num_channels,
const int input_frames,
const int input_height,
const int input_width,
const int output_frames,
const int output_height,
const int output_width,
const float temporal_scale,
const float height_scale,
const float width_scale,
const float* dY,
float* dX) {
CUDA_1D_KERNEL_LOOP(index, size) {
int indexTemp = index;
const int x = indexTemp % input_width;
indexTemp /= input_width;
const int y = indexTemp % input_height;
indexTemp /= input_height;
const int f = indexTemp % input_frames;
indexTemp /= input_frames;
const int c = indexTemp % num_channels;
indexTemp /= num_channels;
const int n = indexTemp;
const int out_f = fminf(f / temporal_scale, output_frames - 1);
const int out_y = fminf(y / height_scale, output_height - 1);
const int out_x = fminf(x / width_scale, output_width - 1);
const int out_index =
(((n * num_channels + c) * output_frames + out_f) * output_height +
out_y) * output_width + out_x;
#if __CUDA_ARCH__ >= 350
gpu_atomic_add(dX + out_index, __ldg(dY + index));
#else
gpu_atomic_add(dX + out_index, *(dY + index));
#endif
}
}
} // namespace
template <>
bool ResizeNearest3DOp<float, CUDAContext>::RunOnDeviceWithOrderNCHW() {
const auto& X = Input(0);
const auto inputDims = X.sizes();
CAFFE_ENFORCE_EQ(5, inputDims.size());
const int batch_size = X.dim32(0), num_channels = X.dim32(1),
input_frames = X.dim32(2), input_height = X.dim32(3),
input_width = X.dim32(4);
CAFFE_ENFORCE_EQ(InputSize(), 1);
int output_frames = input_frames * temporal_scale_;
int output_height = input_height * height_scale_;
int output_width = input_width * width_scale_;
auto* Y = Output(
0,
{batch_size, num_channels, output_frames, output_height, output_width},
at::dtype<float>());
const auto size = Y->numel();
NearestNeighbor3DKernel<<<
CAFFE_GET_BLOCKS(size),
CAFFE_CUDA_NUM_THREADS,
0,
context_.cuda_stream()>>>(
size,
num_channels,
input_frames,
input_height,
input_width,
output_frames,
output_height,
output_width,
temporal_scale_,
height_scale_,
width_scale_,
X.data<float>(),
Y->template mutable_data<float>());
C10_CUDA_KERNEL_LAUNCH_CHECK();
return true;
}
template <>
bool ResizeNearest3DOp<float, CUDAContext>::RunOnDevice() {
switch (order_) {
case StorageOrder::NHWC:
CAFFE_THROW("Not implemented for storage order: ", order_);
case StorageOrder::NCHW:
return RunOnDeviceWithOrderNCHW();
default:
CAFFE_THROW("Unknown Storage order: ", order_);
}
}
template <>
bool ResizeNearest3DGradientOp<float, CUDAContext>::RunOnDeviceWithOrderNCHW() {
const auto& dY = Input(0);
const auto& X = Input(1);
const auto inputDims = dY.sizes();
CAFFE_ENFORCE_EQ(5, inputDims.size());
const int batch_size = dY.dim32(0), num_channels = dY.dim32(1),
input_frames = dY.dim32(2), input_height = dY.dim32(3),
input_width = dY.dim32(4);
// X,dim32(2) can be different from int(input_frames / temporal_scale_)
// We choose to compute output_frames=int(input_frames / temporal_scale_)
// const int output_frames = X,dim32(2);
// const int output_height = X.dim32(3);
// const int output_width = X.dim32(4);
const int output_frames = int(input_frames / temporal_scale_);
const int output_height = int(input_height / height_scale_);
const int output_width = int(input_width / width_scale_);
CAFFE_ENFORCE_EQ(InputSize(), 2);
auto* dX = Output(
0,
{batch_size, num_channels, output_frames, output_height, output_width},
at::dtype<float>());
math::Set<float, CUDAContext>(
dX->numel(), 0.0f, dX->template mutable_data<float>(), &context_);
const auto size = dY.numel();
NearestNeighbor3DGradientKernel<<<
CAFFE_GET_BLOCKS(size),
CAFFE_CUDA_NUM_THREADS,
0,
context_.cuda_stream()>>>(
size,
num_channels,
input_frames,
input_height,
input_width,
output_frames,
output_height,
output_width,
temporal_scale_,
height_scale_,
width_scale_,
dY.data<float>(),
dX->template mutable_data<float>());
C10_CUDA_KERNEL_LAUNCH_CHECK();
return true;
}
template <>
bool ResizeNearest3DGradientOp<float, CUDAContext>::RunOnDevice() {
switch (order_) {
case StorageOrder::NHWC:
CAFFE_THROW("Not implemented for storage order: ", order_);
case StorageOrder::NCHW:
return RunOnDeviceWithOrderNCHW();
default:
CAFFE_THROW("Unknown Storage order: ", order_);
}
}
REGISTER_CUDA_OPERATOR(ResizeNearest3D, ResizeNearest3DOp<float, CUDAContext>);
REGISTER_CUDA_OPERATOR(
ResizeNearest3DGradient,
ResizeNearest3DGradientOp<float, CUDAContext>);
} // namespace caffe2
using ResizeNearest3DOpFloatCUDA =
caffe2::ResizeNearest3DOp<float, caffe2::CUDAContext>;
C10_EXPORT_CAFFE2_OP_TO_C10_CUDA(ResizeNearest3D, ResizeNearest3DOpFloatCUDA);
|
