From ccbab41039af424237eaac5c302c293ab97540f8 Mon Sep 17 00:00:00 2001 From: "Guo, Yejun" Date: Mon, 29 Jul 2019 09:56:54 +0800 Subject: dnn: convert tf.pad to native model in python script, and load/execute it in the c code. since tf.pad is enabled, the conv2d(valid) changes back to its original behavior. Signed-off-by: Guo, Yejun Signed-off-by: Pedro Arthur --- libavfilter/dnn/dnn_backend_native.c | 35 +++++++++++++++++++++++++++++++++++ libavfilter/dnn/dnn_backend_native.h | 2 +- 2 files changed, 36 insertions(+), 1 deletion(-) (limited to 'libavfilter') diff --git a/libavfilter/dnn/dnn_backend_native.c b/libavfilter/dnn/dnn_backend_native.c index 82e900b..09c583b 100644 --- a/libavfilter/dnn/dnn_backend_native.c +++ b/libavfilter/dnn/dnn_backend_native.c @@ -25,6 +25,7 @@ #include "dnn_backend_native.h" #include "libavutil/avassert.h" +#include "dnn_backend_native_layer_pad.h" static DNNReturnType set_input_output_native(void *model, DNNInputData *input, const char *input_name, const char **output_names, uint32_t nb_output) { @@ -32,6 +33,7 @@ static DNNReturnType set_input_output_native(void *model, DNNInputData *input, c InputParams *input_params; ConvolutionalParams *conv_params; DepthToSpaceParams *depth_to_space_params; + LayerPadParams *pad_params; int cur_width, cur_height, cur_channels; int32_t layer; @@ -77,6 +79,12 @@ static DNNReturnType set_input_output_native(void *model, DNNInputData *input, c cur_height *= depth_to_space_params->block_size; cur_width *= depth_to_space_params->block_size; break; + case MIRROR_PAD: + pad_params = (LayerPadParams *)network->layers[layer].params; + cur_height = cur_height + pad_params->paddings[1][0] + pad_params->paddings[1][1]; + cur_width = cur_width + pad_params->paddings[2][0] + pad_params->paddings[2][1]; + cur_channels = cur_channels + pad_params->paddings[3][0] + pad_params->paddings[3][1]; + break; default: return DNN_ERROR; } @@ -110,6 +118,7 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) DNNLayerType layer_type; ConvolutionalParams *conv_params; DepthToSpaceParams *depth_to_space_params; + LayerPadParams *pad_params; model = av_malloc(sizeof(DNNModel)); if (!model){ @@ -207,6 +216,23 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) network->layers[layer].type = DEPTH_TO_SPACE; network->layers[layer].params = depth_to_space_params; break; + case MIRROR_PAD: + pad_params = av_malloc(sizeof(LayerPadParams)); + if (!pad_params){ + avio_closep(&model_file_context); + ff_dnn_free_model_native(&model); + return NULL; + } + pad_params->mode = (int32_t)avio_rl32(model_file_context); + dnn_size += 4; + for (i = 0; i < 4; ++i) { + pad_params->paddings[i][0] = avio_rl32(model_file_context); + pad_params->paddings[i][1] = avio_rl32(model_file_context); + dnn_size += 8; + } + network->layers[layer].type = MIRROR_PAD; + network->layers[layer].params = pad_params; + break; default: avio_closep(&model_file_context); ff_dnn_free_model_native(&model); @@ -314,6 +340,7 @@ DNNReturnType ff_dnn_execute_model_native(const DNNModel *model, DNNData *output InputParams *input_params; ConvolutionalParams *conv_params; DepthToSpaceParams *depth_to_space_params; + LayerPadParams *pad_params; if (network->layers_num <= 0 || network->layers[0].type != INPUT || !network->layers[0].output){ return DNN_ERROR; @@ -348,6 +375,14 @@ DNNReturnType ff_dnn_execute_model_native(const DNNModel *model, DNNData *output cur_width *= depth_to_space_params->block_size; cur_channels /= depth_to_space_params->block_size * depth_to_space_params->block_size; break; + case MIRROR_PAD: + pad_params = (LayerPadParams *)network->layers[layer].params; + dnn_execute_layer_pad(network->layers[layer - 1].output, network->layers[layer].output, + pad_params, 1, cur_height, cur_width, cur_channels); + cur_height = cur_height + pad_params->paddings[1][0] + pad_params->paddings[1][1]; + cur_width = cur_width + pad_params->paddings[2][0] + pad_params->paddings[2][1]; + cur_channels = cur_channels + pad_params->paddings[3][0] + pad_params->paddings[3][1]; + break; case INPUT: return DNN_ERROR; } diff --git a/libavfilter/dnn/dnn_backend_native.h b/libavfilter/dnn/dnn_backend_native.h index 8ef1855..b6f9533 100644 --- a/libavfilter/dnn/dnn_backend_native.h +++ b/libavfilter/dnn/dnn_backend_native.h @@ -30,7 +30,7 @@ #include "../dnn_interface.h" #include "libavformat/avio.h" -typedef enum {INPUT, CONV, DEPTH_TO_SPACE} DNNLayerType; +typedef enum {INPUT, CONV, DEPTH_TO_SPACE, MIRROR_PAD} DNNLayerType; typedef enum {RELU, TANH, SIGMOID, NONE, LEAKY_RELU} DNNActivationFunc; -- cgit v1.1