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
|
#include <vector>
#include "caffe/common.hpp"
#include "caffe/layer.hpp"
#include "caffe/util/im2col.hpp"
#include "caffe/vision_layers.hpp"
namespace caffe {
template <typename Dtype>
void Im2colLayer<Dtype>::LayerSetUp(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
ConvolutionParameter conv_param = this->layer_param_.convolution_param();
force_nd_im2col_ = conv_param.force_nd_im2col();
const int input_num_dims = bottom[0]->shape().size();
channel_axis_ = bottom[0]->CanonicalAxisIndex(conv_param.axis());
const int first_spatial_dim = channel_axis_ + 1;
num_spatial_axes_ = input_num_dims - first_spatial_dim;
CHECK_GE(num_spatial_axes_, 1);
vector<int> dim_blob_shape(1, num_spatial_axes_);
// Setup filter kernel dimensions (kernel_shape_).
kernel_shape_.Reshape(dim_blob_shape);
int* kernel_shape_data = kernel_shape_.mutable_cpu_data();
if (conv_param.has_kernel_h() || conv_param.has_kernel_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "kernel_h & kernel_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.kernel_size_size())
<< "Either kernel_size or kernel_h/w should be specified; not both.";
kernel_shape_data[0] = conv_param.kernel_h();
kernel_shape_data[1] = conv_param.kernel_w();
} else {
const int num_kernel_dims = conv_param.kernel_size_size();
CHECK(num_kernel_dims == 1 || num_kernel_dims == num_spatial_axes_)
<< "kernel_size must be specified once, or once per spatial dimension "
<< "(kernel_size specified " << num_kernel_dims << " times; "
<< num_spatial_axes_ << " spatial dims);";
for (int i = 0; i < num_spatial_axes_; ++i) {
kernel_shape_data[i] =
conv_param.kernel_size((num_kernel_dims == 1) ? 0 : i);
}
}
for (int i = 0; i < num_spatial_axes_; ++i) {
CHECK_GT(kernel_shape_data[i], 0) << "Filter dimensions must be nonzero.";
}
// Setup stride dimensions (stride_).
stride_.Reshape(dim_blob_shape);
int* stride_data = stride_.mutable_cpu_data();
if (conv_param.has_stride_h() || conv_param.has_stride_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "stride_h & stride_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.stride_size())
<< "Either stride or stride_h/w should be specified; not both.";
stride_data[0] = conv_param.stride_h();
stride_data[1] = conv_param.stride_w();
} else {
const int num_stride_dims = conv_param.stride_size();
CHECK(num_stride_dims == 0 || num_stride_dims == 1 ||
num_stride_dims == num_spatial_axes_)
<< "stride must be specified once, or once per spatial dimension "
<< "(stride specified " << num_stride_dims << " times; "
<< num_spatial_axes_ << " spatial dims);";
const int kDefaultStride = 1;
for (int i = 0; i < num_spatial_axes_; ++i) {
stride_data[i] = (num_stride_dims == 0) ? kDefaultStride :
conv_param.stride((num_stride_dims == 1) ? 0 : i);
CHECK_GT(stride_data[i], 0) << "Stride dimensions must be nonzero.";
}
}
// Setup pad dimensions (pad_).
pad_.Reshape(dim_blob_shape);
int* pad_data = pad_.mutable_cpu_data();
if (conv_param.has_pad_h() || conv_param.has_pad_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "pad_h & pad_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.pad_size())
<< "Either pad or pad_h/w should be specified; not both.";
pad_data[0] = conv_param.pad_h();
pad_data[1] = conv_param.pad_w();
} else {
const int num_pad_dims = conv_param.pad_size();
CHECK(num_pad_dims == 0 || num_pad_dims == 1 ||
num_pad_dims == num_spatial_axes_)
<< "pad must be specified once, or once per spatial dimension "
<< "(pad specified " << num_pad_dims << " times; "
<< num_spatial_axes_ << " spatial dims);";
const int kDefaultPad = 0;
for (int i = 0; i < num_spatial_axes_; ++i) {
pad_data[i] = (num_pad_dims == 0) ? kDefaultPad :
conv_param.pad((num_pad_dims == 1) ? 0 : i);
}
}
}
template <typename Dtype>
void Im2colLayer<Dtype>::Reshape(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
vector<int> top_shape = bottom[0]->shape();
const int* kernel_shape_data = kernel_shape_.cpu_data();
const int* stride_data = stride_.cpu_data();
const int* pad_data = pad_.cpu_data();
for (int i = 0; i < num_spatial_axes_; ++i) {
top_shape[channel_axis_] *= kernel_shape_data[i];
const int input_dim = bottom[0]->shape(channel_axis_ + i + 1);
const int output_dim = (input_dim + 2 * pad_data[i] - kernel_shape_data[i])
/ stride_data[i] + 1;
top_shape[channel_axis_ + i + 1] = output_dim;
}
top[0]->Reshape(top_shape);
num_ = bottom[0]->count(0, channel_axis_);
bottom_dim_ = bottom[0]->count(channel_axis_);
top_dim_ = top[0]->count(channel_axis_);
channels_ = bottom[0]->shape(channel_axis_);
}
template <typename Dtype>
void Im2colLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
const Dtype* bottom_data = bottom[0]->cpu_data();
Dtype* top_data = top[0]->mutable_cpu_data();
for (int n = 0; n < num_; ++n) {
DCHECK_EQ(bottom[0]->shape().size() - channel_axis_, num_spatial_axes_ + 1);
DCHECK_EQ(top[0]->shape().size() - channel_axis_, num_spatial_axes_ + 1);
DCHECK_EQ(kernel_shape_.count(), num_spatial_axes_);
DCHECK_EQ(pad_.count(), num_spatial_axes_);
DCHECK_EQ(stride_.count(), num_spatial_axes_);
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
im2col_cpu(bottom_data + n * bottom_dim_, channels_,
bottom[0]->shape(channel_axis_ + 1),
bottom[0]->shape(channel_axis_ + 2),
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
top_data + n * top_dim_);
} else {
im2col_nd_cpu(bottom_data + n * bottom_dim_, num_spatial_axes_,
bottom[0]->shape().data() + channel_axis_,
top[0]->shape().data() + channel_axis_,
kernel_shape_.cpu_data(), pad_.cpu_data(), stride_.cpu_data(),
top_data + n * top_dim_);
}
}
}
template <typename Dtype>
void Im2colLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom) {
const Dtype* top_diff = top[0]->cpu_diff();
Dtype* bottom_diff = bottom[0]->mutable_cpu_diff();
for (int n = 0; n < num_; ++n) {
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
col2im_cpu(top_diff + n * top_dim_, channels_,
bottom[0]->shape(channel_axis_ + 1),
bottom[0]->shape(channel_axis_ + 2),
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
bottom_diff + n * bottom_dim_);
} else {
col2im_nd_cpu(top_diff + n * top_dim_, num_spatial_axes_,
bottom[0]->shape().data() + channel_axis_,
top[0]->shape().data() + channel_axis_,
kernel_shape_.cpu_data(), pad_.cpu_data(), stride_.cpu_data(),
bottom_diff + n * bottom_dim_);
}
}
}
#ifdef CPU_ONLY
STUB_GPU(Im2colLayer);
#endif
INSTANTIATE_CLASS(Im2colLayer);
REGISTER_LAYER_CLASS(Im2col);
} // namespace caffe
|
