Chiebot-Mirror
/
opencv
8
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

ocl4dnn: Fix SAME padding mode for convolve

Browse Source 
Signed-off-by: Wu, Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Li Peng <peng.li@intel.com>
pull/10955/head
Wu Zhiwen 7 years ago committed by Li Peng
parent 24bed38c2b
commit ef937dd676
5 changed files with 38 additions and 28 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 3
      modules/dnn/src/layers/convolution_layer.cpp
  2. 2
      modules/dnn/src/ocl4dnn/include/ocl4dnn.hpp
  3. 8
      modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
  4. 51
      modules/dnn/src/opencl/conv_layer_spatial.cl
  5. 2
      modules/dnn/test/test_tf_importer.cpp

3
modules/dnn/src/layers/convolution_layer.cpp
Unescape View File

@ -824,9 +824,6 @@ public:
for (int i = 0; i < inputs.size(); ++i) for (int i = 0; i < inputs.size(); ++i)
CV_Assert(inputs[i].u != outputs[0].u); CV_Assert(inputs[i].u != outputs[0].u);
if (padMode == "SAME")
return false;
if (convolutionOp.empty()) if (convolutionOp.empty())
{ {
OCL4DNNConvConfig config; OCL4DNNConvConfig config;

2
modules/dnn/src/ocl4dnn/include/ocl4dnn.hpp
Unescape View File

@ -285,6 +285,8 @@ class OCL4DNNConvSpatial
int32_t width_; int32_t width_;
int32_t pad_h_; int32_t pad_h_;
int32_t pad_w_; int32_t pad_w_;
int32_t pad_bottom_;
int32_t pad_right_;
int32_t stride_h_; int32_t stride_h_;
int32_t stride_w_; int32_t stride_w_;
int32_t dilation_h_; int32_t dilation_h_;

8
modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
Unescape View File

@ -103,6 +103,12 @@ OCL4DNNConvSpatial<Dtype>::OCL4DNNConvSpatial(OCL4DNNConvConfig config)
output_w_ = config.out_shape[dims - spatial_dims + 1]; output_w_ = config.out_shape[dims - spatial_dims + 1];
bottom_dim_ = channels_ * width_ * height_; bottom_dim_ = channels_ * width_ * height_;
top_dim_ = num_output_ * output_w_ * output_h_; top_dim_ = num_output_ * output_w_ * output_h_;
int Ph = (output_h_ - 1) * stride_h_ + (dilation_h_ * (kernel_h_ - 1) + 1) - height_;
int Pw = (output_w_ - 1) * stride_w_ + (dilation_w_ * (kernel_w_ - 1) + 1) - width_;
Ph = (Ph > 0) ? Ph : 0;
Pw = (Pw > 0) ? Pw : 0;
pad_right_ = (Pw + 1) / 2;
pad_bottom_ = (Ph + 1) / 2;
cache_path_ = utils::getConfigurationParameterString("OPENCV_OCL4DNN_CONFIG_PATH", ""); cache_path_ = utils::getConfigurationParameterString("OPENCV_OCL4DNN_CONFIG_PATH", "");
dwconv_ = (num_output_ == channels_ && channels_ == group_); dwconv_ = (num_output_ == channels_ && channels_ == group_);
@ -379,6 +385,8 @@ void OCL4DNNConvSpatial<Dtype>::setupKernel()
{ {
addDef("INPUT_PAD_W", pad_w_); addDef("INPUT_PAD_W", pad_w_);
addDef("INPUT_PAD_H", pad_h_); addDef("INPUT_PAD_H", pad_h_);
addDef("INPUT_PAD_RIGHT", pad_right_);
addDef("INPUT_PAD_BOTTOM", pad_bottom_);
} }
setupKernelDetails(kernelType_, blockM_, blockK_, blockN_); setupKernelDetails(kernelType_, blockM_, blockK_, blockN_);

51
modules/dnn/src/opencl/conv_layer_spatial.cl
Unescape View File

@ -238,7 +238,7 @@ convolve_simd(
int curr_local_x = ( lid % ( TILE_X / 4 ) ) * 4; int curr_local_x = ( lid % ( TILE_X / 4 ) ) * 4;
int curr_y = or * STRIDE_Y + curr_local_y; int curr_y = or * STRIDE_Y + curr_local_y;
int curr_x = oc * STRIDE_X + curr_local_x; int curr_x = oc * STRIDE_X + curr_local_x;
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
in_addr = input_batch_offset in_addr = input_batch_offset
@ -256,19 +256,22 @@ convolve_simd(
LOOP(INVEC_SIZE, reg, LOOP(INVEC_SIZE, reg,
{ {
if (curr_local_y + reg * TILE_Y_STRIDE < TILE_Y || INVEC_SIZE * TILE_Y_STRIDE <= (TILE_Y + 2) || reg < INVEC_SIZE - 1) { if (curr_local_y + reg * TILE_Y_STRIDE < TILE_Y || INVEC_SIZE * TILE_Y_STRIDE <= (TILE_Y + 2) || reg < INVEC_SIZE - 1) {
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
if (curr_y >= INPUT_PAD_H && curr_y < input_height + INPUT_PAD_H && curr_x + 3 >= INPUT_PAD_W && curr_x < input_width + INPUT_PAD_W) { if (curr_y >= INPUT_PAD_H && curr_y < input_height + INPUT_PAD_H && curr_x + 3 >= INPUT_PAD_W && curr_x < input_width + INPUT_PAD_W) {
if (curr_x < INPUT_PAD_W) { if (curr_x < INPUT_PAD_W) {
in_buf.in_vec[reg].s0 = 0; in_buf.in_vec[reg].s0 = 0;
if (curr_x + 1 >= INPUT_PAD_W) if (curr_x + 1 >= INPUT_PAD_W && curr_x + 1 < input_width + INPUT_PAD_W)
in_buf.in_vec[reg].s1 = *(inputs + in_offset + 1); in_buf.in_vec[reg].s1 = *(inputs + in_offset + 1);
else else
in_buf.in_vec[reg].s1 = 0; in_buf.in_vec[reg].s1 = 0;
if (curr_x + 2 >= INPUT_PAD_W) if (curr_x + 2 >= INPUT_PAD_W && curr_x + 2 < input_width + INPUT_PAD_W)
in_buf.in_vec[reg].s2 = *(inputs + in_offset + 2); in_buf.in_vec[reg].s2 = *(inputs + in_offset + 2);
else else
in_buf.in_vec[reg].s2 = 0; in_buf.in_vec[reg].s2 = 0;
in_buf.in_vec[reg].s3 = *(inputs + in_offset + 3); if (curr_x + 3 < input_width + INPUT_PAD_W)
in_buf.in_vec[reg].s3 = *(inputs + in_offset + 3);
else
in_buf.in_vec[reg].s3 = 0;
} else { } else {
VLOAD4(in_buf.in_vec[reg], inputs + in_offset); VLOAD4(in_buf.in_vec[reg], inputs + in_offset);
if (curr_x + 1 >= input_width + INPUT_PAD_W) if (curr_x + 1 >= input_width + INPUT_PAD_W)
@ -289,7 +292,7 @@ convolve_simd(
in_offset += input_width * TILE_Y_STRIDE; in_offset += input_width * TILE_Y_STRIDE;
}); });
in_addr += input_height * input_width; in_addr += input_height * input_width;
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y = saved_y; curr_y = saved_y;
#endif #endif
@ -492,7 +495,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// atile is M rows x K columns. // atile is M rows x K columns.
int curr_x = ( global_y % output_width ) * STRIDE_X; int curr_x = ( global_y % output_width ) * STRIDE_X;
int curr_y = ( global_y / output_width ) * STRIDE_Y; int curr_y = ( global_y / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
const __global Dtype *src0_read = src0 const __global Dtype *src0_read = src0
@ -512,7 +515,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
do do
{ {
int patch_row = 0; int patch_row = 0;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y = saved_y; curr_y = saved_y;
#endif #endif
@ -530,7 +533,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// ... // ...
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ]; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ];
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#else #else
@ -646,7 +649,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// atile is M rows x K columns. // atile is M rows x K columns.
int curr_x = ( global_y % output_width ) * STRIDE_X; int curr_x = ( global_y % output_width ) * STRIDE_X;
int curr_y = ( global_y / output_width ) * STRIDE_Y; int curr_y = ( global_y / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
const __global Dtype *src0_read = src0 const __global Dtype *src0_read = src0
@ -666,14 +669,14 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
do do
{ {
int patch_row = 0; int patch_row = 0;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y = saved_y; curr_y = saved_y;
#endif #endif
do do
{ {
// Load atile and interleaved btile. // Load atile and interleaved btile.
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ]; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ];
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#else #else
@ -873,7 +876,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X; int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X;
int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y; int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y;
int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y; int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y0 = curr_y0; int saved_y0 = curr_y0;
int saved_y1 = curr_y1; int saved_y1 = curr_y1;
#endif #endif
@ -911,7 +914,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// (0, 2) (8, 2) (16, 2) (24, 2) ... ... // (0, 2) (8, 2) (16, 2) (24, 2) ... ...
// ... // ...
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH;
Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH; Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH;
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
@ -997,7 +1000,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
//while( ++patch_row < 1 ); //debug //while( ++patch_row < 1 ); //debug
while( ++patch_row < KERNEL_HEIGHT ); while( ++patch_row < KERNEL_HEIGHT );
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y0 = saved_y0; curr_y0 = saved_y0;
curr_y1 = saved_y1; curr_y1 = saved_y1;
#endif #endif
@ -1073,7 +1076,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X; int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X;
int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y; int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y;
int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y; int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y0 = curr_y0; int saved_y0 = curr_y0;
int saved_y1 = curr_y1; int saved_y1 = curr_y1;
#endif #endif
@ -1102,7 +1105,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
{ {
// Load atile and interleaved btile. // Load atile and interleaved btile.
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH;
Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH; Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH;
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
@ -1210,7 +1213,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
//while( ++patch_row < 1 ); //debug //while( ++patch_row < 1 ); //debug
while( ++patch_row < KERNEL_HEIGHT ); while( ++patch_row < KERNEL_HEIGHT );
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y0 = saved_y0; curr_y0 = saved_y0;
curr_y1 = saved_y1; curr_y1 = saved_y1;
#endif #endif
@ -1377,7 +1380,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// atile is M rows x K columns. // atile is M rows x K columns.
int curr_x = ( global_y % output_width ) * STRIDE_X; int curr_x = ( global_y % output_width ) * STRIDE_X;
int curr_y = ( global_y / output_width ) * STRIDE_Y; int curr_y = ( global_y / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
const __global Dtype *src0_read = src0 const __global Dtype *src0_read = src0
@ -1419,7 +1422,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
do do
{ {
int patch_row = 0; int patch_row = 0;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y = saved_y; curr_y = saved_y;
#endif #endif
__attribute__((opencl_unroll_hint(1))) __attribute__((opencl_unroll_hint(1)))
@ -1437,7 +1440,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// ... // ...
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ]; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ];
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#else #else
@ -1580,7 +1583,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X; int curr_x1 = ( ( global_y * TILE_M + 1 ) % output_width ) * STRIDE_X;
int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y; int curr_y0 = ( ( global_y * TILE_M + 0 ) / output_width ) * STRIDE_Y;
int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y; int curr_y1 = ( ( global_y * TILE_M + 1 ) / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
int saved_y0 = curr_y0; int saved_y0 = curr_y0;
int saved_y1 = curr_y1; int saved_y1 = curr_y1;
#endif #endif
@ -1618,7 +1621,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// (0, 2) (8, 2) (16, 2) (24, 2) ... ... // (0, 2) (8, 2) (16, 2) (24, 2) ... ...
// ... // ...
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 #if INPUT_PAD_H == 0 && INPUT_PAD_W == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read0 )[ 0 ]; src0_read0 += ROW_PITCH;
Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH; Dtype_t blockA01 = ( (const __global Dtype_t*)src0_read1 )[ 0 ]; src0_read1 += ROW_PITCH;
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
@ -1692,7 +1695,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
//while( ++patch_row < 1 ); //debug //while( ++patch_row < 1 ); //debug
while( ++patch_row < KERNEL_HEIGHT ); while( ++patch_row < KERNEL_HEIGHT );
#if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 #if INPUT_PAD_W != 0 || INPUT_PAD_H != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0
curr_y0 = saved_y0; curr_y0 = saved_y0;
curr_y1 = saved_y1; curr_y1 = saved_y1;
#endif #endif

2
modules/dnn/test/test_tf_importer.cpp
Unescape View File

@ -321,7 +321,7 @@ OCL_TEST(Test_TensorFlow, MobileNet_SSD)
std::vector<Mat> output; std::vector<Mat> output;
net.forward(output, outNames); net.forward(output, outNames);
normAssert(target[0].reshape(1, 1), output[0].reshape(1, 1)); normAssert(target[0].reshape(1, 1), output[0].reshape(1, 1), "", 1e-5, 1.5e-4);
normAssert(target[1].reshape(1, 1), output[1].reshape(1, 1), "", 1e-5, 3e-4); normAssert(target[1].reshape(1, 1), output[1].reshape(1, 1), "", 1e-5, 3e-4);
normAssert(target[2].reshape(1, 1), output[2].reshape(1, 1), "", 4e-5, 1e-2); normAssert(target[2].reshape(1, 1), output[2].reshape(1, 1), "", 4e-5, 1e-2);
} }

Write Preview
Loading…
Cancel
Save