| @@ -188,8 +188,9 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) · 187 unmodified lines |
| 188 | 188 | for (layer = 0; layer < network->layers_num; ++layer){ |
| 189 | 189 | layer_type = (int32_t)avio_rl32(model_file_context); |
| 190 | 190 | dnn_size += 4; |
| 191 | + network->layers[layer].type = layer_type; |
| 191 | 192 | switch (layer_type){ |
| 192 | | − case CONV: |
| 193 | + case DLT_CONV2D: |
| 193 | 194 | conv_params = av_malloc(sizeof(ConvolutionalParams)); |
| 194 | 195 | if (!conv_params){ |
| 195 | 196 | avio_closep(&model_file_context); |
| @@ -231,10 +232,9 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) · 35 unmodified lines |
| 231 | 232 | network->layers[layer].input_operand_indexes[0] = (int32_t)avio_rl32(model_file_context); |
| 232 | 233 | network->layers[layer].output_operand_index = (int32_t)avio_rl32(model_file_context); |
| 233 | 234 | dnn_size += 8; |
| 234 | | − network->layers[layer].type = CONV; |
| 235 | 235 | network->layers[layer].params = conv_params; |
| 236 | 236 | break; |
| 237 | | − case DEPTH_TO_SPACE: |
| 237 | + case DLT_DEPTH_TO_SPACE: |
| 238 | 238 | depth_to_space_params = av_malloc(sizeof(DepthToSpaceParams)); |
| 239 | 239 | if (!depth_to_space_params){ |
| 240 | 240 | avio_closep(&model_file_context); |
| @@ -246,10 +246,9 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) · 5 unmodified lines |
| 246 | 246 | network->layers[layer].input_operand_indexes[0] = (int32_t)avio_rl32(model_file_context); |
| 247 | 247 | network->layers[layer].output_operand_index = (int32_t)avio_rl32(model_file_context); |
| 248 | 248 | dnn_size += 8; |
| 249 | | − network->layers[layer].type = DEPTH_TO_SPACE; |
| 250 | 249 | network->layers[layer].params = depth_to_space_params; |
| 251 | 250 | break; |
| 252 | | − case MIRROR_PAD: |
| 251 | + case DLT_MIRROR_PAD: |
| 253 | 252 | pad_params = av_malloc(sizeof(LayerPadParams)); |
| 254 | 253 | if (!pad_params){ |
| 255 | 254 | avio_closep(&model_file_context); |
| @@ -266,10 +265,9 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) · 10 unmodified lines |
| 266 | 265 | network->layers[layer].input_operand_indexes[0] = (int32_t)avio_rl32(model_file_context); |
| 267 | 266 | network->layers[layer].output_operand_index = (int32_t)avio_rl32(model_file_context); |
| 268 | 267 | dnn_size += 8; |
| 269 | | − network->layers[layer].type = MIRROR_PAD; |
| 270 | 268 | network->layers[layer].params = pad_params; |
| 271 | 269 | break; |
| 272 | | − case MAXIMUM: |
| 270 | + case DLT_MAXIMUM: |
| 273 | 271 | maximum_params = av_malloc(sizeof(*maximum_params)); |
| 274 | 272 | if (!maximum_params){ |
| 275 | 273 | avio_closep(&model_file_context); |
| @@ -278,7 +276,6 @@ DNNModel *ff_dnn_load_model_native(const char *model_filename) · 2 unmodified lines |
| 278 | 276 | } |
| 279 | 277 | maximum_params->val.u32 = avio_rl32(model_file_context); |
| 280 | 278 | dnn_size += 4; |
| 281 | | − network->layers[layer].type = MAXIMUM; |
| 282 | 279 | network->layers[layer].params = maximum_params; |
| 283 | 280 | network->layers[layer].input_operand_indexes[0] = (int32_t)avio_rl32(model_file_context); |
| 284 | 281 | network->layers[layer].output_operand_index = (int32_t)avio_rl32(model_file_context); |
| @@ -347,27 +344,27 @@ DNNReturnType ff_dnn_execute_model_native(const DNNModel *model, DNNData *output · 62 unmodified lines |
| 347 | 344 | |
| 348 | 345 | for (layer = 0; layer < network->layers_num; ++layer){ |
| 349 | 346 | switch (network->layers[layer].type){ |
| 350 | | − case CONV: |
| 347 | + case DLT_CONV2D: |
| 351 | 348 | conv_params = (ConvolutionalParams *)network->layers[layer].params; |
| 352 | 349 | convolve(network->operands, network->layers[layer].input_operand_indexes, |
| 353 | 350 | network->layers[layer].output_operand_index, conv_params); |
| 354 | 351 | break; |
| 355 | | − case DEPTH_TO_SPACE: |
| 352 | + case DLT_DEPTH_TO_SPACE: |
| 356 | 353 | depth_to_space_params = (DepthToSpaceParams *)network->layers[layer].params; |
| 357 | 354 | depth_to_space(network->operands, network->layers[layer].input_operand_indexes, |
| 358 | 355 | network->layers[layer].output_operand_index, depth_to_space_params->block_size); |
| 359 | 356 | break; |
| 360 | | − case MIRROR_PAD: |
| 357 | + case DLT_MIRROR_PAD: |
| 361 | 358 | pad_params = (LayerPadParams *)network->layers[layer].params; |
| 362 | 359 | dnn_execute_layer_pad(network->operands, network->layers[layer].input_operand_indexes, |
| 363 | 360 | network->layers[layer].output_operand_index, pad_params); |
| 364 | 361 | break; |
| 365 | | − case MAXIMUM: |
| 362 | + case DLT_MAXIMUM: |
| 366 | 363 | maximum_params = (DnnLayerMaximumParams *)network->layers[layer].params; |
| 367 | 364 | dnn_execute_layer_maximum(network->operands, network->layers[layer].input_operand_indexes, |
| 368 | 365 | network->layers[layer].output_operand_index, maximum_params); |
| 369 | 366 | break; |
| 370 | | − case INPUT: |
| 367 | + case DLT_INPUT: |
| 371 | 368 | return DNN_ERROR; |
| 372 | 369 | } |
| 373 | 370 | } |
| @@ -408,7 +405,7 @@ void ff_dnn_free_model_native(DNNModel **model) · 34 unmodified lines |
| 408 | 405 | { |
| 409 | 406 | network = (ConvolutionalNetwork *)(*model)->model; |
| 410 | 407 | for (layer = 0; layer < network->layers_num; ++layer){ |
| 411 | | − if (network->layers[layer].type == CONV){ |
| 408 | + if (network->layers[layer].type == DLT_CONV2D){ |
| 412 | 409 | conv_params = (ConvolutionalParams *)network->layers[layer].params; |
| 413 | 410 | av_freep(&conv_params->kernel); |
| 414 | 411 | av_freep(&conv_params->biases); |