libavfilter/dnn: move dnn files from libavfilter to libavfilter/dnn

it is expected that there will be more files to support native mode,
so put all the dnn codes under libavfilter/dnn

The main change of this patch is to move the file location, see below:
modified:   libavfilter/Makefile
new file:   libavfilter/dnn/Makefile
renamed:    libavfilter/dnn_backend_native.c -> libavfilter/dnn/dnn_backend_native.c
renamed:    libavfilter/dnn_backend_native.h -> libavfilter/dnn/dnn_backend_native.h
renamed:    libavfilter/dnn_backend_tf.c -> libavfilter/dnn/dnn_backend_tf.c
renamed:    libavfilter/dnn_backend_tf.h -> libavfilter/dnn/dnn_backend_tf.h
renamed:    libavfilter/dnn_interface.c -> libavfilter/dnn/dnn_interface.c

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>

6 files changed, 1173 insertions, 0 deletions

diff --git a/libavfilter/dnn/Makefile b/libavfilter/dnn/Makefile
new file mode 100644
index 0000000000..1d12ade165
--- /dev/null
+++ b/libavfilter/dnn/Makefile
@@ -0,0 +1,6 @@
+OBJS-$(CONFIG_DNN)                           += dnn/dnn_interface.o
+OBJS-$(CONFIG_DNN)                           += dnn/dnn_backend_native.o
+
+DNN-OBJS-$(CONFIG_LIBTENSORFLOW)             += dnn/dnn_backend_tf.o
+
+OBJS-$(CONFIG_DNN)                           += $(DNN-OBJS-yes)
diff --git a/libavfilter/dnn/dnn_backend_native.c b/libavfilter/dnn/dnn_backend_native.c
new file mode 100644
index 0000000000..82e900bd8c
--- /dev/null
+++ b/libavfilter/dnn/dnn_backend_native.c
@@ -0,0 +1,389 @@
+/*
+ * Copyright (c) 2018 Sergey Lavrushkin
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * DNN native backend implementation.
+ */
+
+#include "dnn_backend_native.h"
+#include "libavutil/avassert.h"
+
+static DNNReturnType set_input_output_native(void *model, DNNInputData *input, const char *input_name, const char **output_names, uint32_t nb_output)
+{
+    ConvolutionalNetwork *network = (ConvolutionalNetwork *)model;
+    InputParams *input_params;
+    ConvolutionalParams *conv_params;
+    DepthToSpaceParams *depth_to_space_params;
+    int cur_width, cur_height, cur_channels;
+    int32_t layer;
+
+    if (network->layers_num <= 0 || network->layers[0].type != INPUT){
+        return DNN_ERROR;
+    }
+    else{
+        input_params = (InputParams *)network->layers[0].params;
+        input_params->width = cur_width = input->width;
+        input_params->height = cur_height = input->height;
+        input_params->channels = cur_channels = input->channels;
+        if (input->data){
+            av_freep(&input->data);
+        }
+        av_assert0(input->dt == DNN_FLOAT);
+        network->layers[0].output = input->data = av_malloc(cur_height * cur_width * cur_channels * sizeof(float));
+        if (!network->layers[0].output){
+            return DNN_ERROR;
+        }
+    }
+
+    for (layer = 1; layer < network->layers_num; ++layer){
+        switch (network->layers[layer].type){
+        case CONV:
+            conv_params = (ConvolutionalParams *)network->layers[layer].params;
+            if (conv_params->input_num != cur_channels){
+                return DNN_ERROR;
+            }
+            cur_channels = conv_params->output_num;
+
+            if (conv_params->padding_method == VALID) {
+                int pad_size = (conv_params->kernel_size - 1) * conv_params->dilation;
+                cur_height -= pad_size;
+                cur_width -= pad_size;
+            }
+            break;
+        case DEPTH_TO_SPACE:
+            depth_to_space_params = (DepthToSpaceParams *)network->layers[layer].params;
+            if (cur_channels % (depth_to_space_params->block_size * depth_to_space_params->block_size) != 0){
+                return DNN_ERROR;
+            }
+            cur_channels = cur_channels / (depth_to_space_params->block_size * depth_to_space_params->block_size);
+            cur_height *= depth_to_space_params->block_size;
+            cur_width *= depth_to_space_params->block_size;
+            break;
+        default:
+            return DNN_ERROR;
+        }
+        if (network->layers[layer].output){
+            av_freep(&network->layers[layer].output);
+        }
+
+        if (cur_height <= 0 || cur_width <= 0)
+            return DNN_ERROR;
+
+        network->layers[layer].output = av_malloc(cur_height * cur_width * cur_channels * sizeof(float));
+        if (!network->layers[layer].output){
+            return DNN_ERROR;
+        }
+    }
+
+    return DNN_SUCCESS;
+}
+
+// Loads model and its parameters that are stored in a binary file with following structure:
+// layers_num,layer_type,layer_parameterss,layer_type,layer_parameters...
+// For CONV layer: activation_function, input_num, output_num, kernel_size, kernel, biases
+// For DEPTH_TO_SPACE layer: block_size
+DNNModel *ff_dnn_load_model_native(const char *model_filename)
+{
+    DNNModel *model = NULL;
+    ConvolutionalNetwork *network = NULL;
+    AVIOContext *model_file_context;
+    int file_size, dnn_size, kernel_size, i;
+    int32_t layer;
+    DNNLayerType layer_type;
+    ConvolutionalParams *conv_params;
+    DepthToSpaceParams *depth_to_space_params;
+
+    model = av_malloc(sizeof(DNNModel));
+    if (!model){
+        return NULL;
+    }
+
+    if (avio_open(&model_file_context, model_filename, AVIO_FLAG_READ) < 0){
+        av_freep(&model);
+        return NULL;
+    }
+    file_size = avio_size(model_file_context);
+
+    network = av_malloc(sizeof(ConvolutionalNetwork));
+    if (!network){
+        avio_closep(&model_file_context);
+        av_freep(&model);
+        return NULL;
+    }
+    model->model = (void *)network;
+
+    network->layers_num = 1 + (int32_t)avio_rl32(model_file_context);
+    dnn_size = 4;
+
+    network->layers = av_malloc(network->layers_num * sizeof(Layer));
+    if (!network->layers){
+        av_freep(&network);
+        avio_closep(&model_file_context);
+        av_freep(&model);
+        return NULL;
+    }
+
+    for (layer = 0; layer < network->layers_num; ++layer){
+        network->layers[layer].output = NULL;
+        network->layers[layer].params = NULL;
+    }
+    network->layers[0].type = INPUT;
+    network->layers[0].params = av_malloc(sizeof(InputParams));
+    if (!network->layers[0].params){
+        avio_closep(&model_file_context);
+        ff_dnn_free_model_native(&model);
+        return NULL;
+    }
+
+    for (layer = 1; layer < network->layers_num; ++layer){
+        layer_type = (int32_t)avio_rl32(model_file_context);
+        dnn_size += 4;
+        switch (layer_type){
+        case CONV:
+            conv_params = av_malloc(sizeof(ConvolutionalParams));
+            if (!conv_params){
+                avio_closep(&model_file_context);
+                ff_dnn_free_model_native(&model);
+                return NULL;
+            }
+            conv_params->dilation = (int32_t)avio_rl32(model_file_context);
+            conv_params->padding_method = (int32_t)avio_rl32(model_file_context);
+            conv_params->activation = (int32_t)avio_rl32(model_file_context);
+            conv_params->input_num = (int32_t)avio_rl32(model_file_context);
+            conv_params->output_num = (int32_t)avio_rl32(model_file_context);
+            conv_params->kernel_size = (int32_t)avio_rl32(model_file_context);
+            kernel_size = conv_params->input_num * conv_params->output_num *
+                          conv_params->kernel_size * conv_params->kernel_size;
+            dnn_size += 24 + (kernel_size + conv_params->output_num << 2);
+            if (dnn_size > file_size || conv_params->input_num <= 0 ||
+                conv_params->output_num <= 0 || conv_params->kernel_size <= 0){
+                avio_closep(&model_file_context);
+                ff_dnn_free_model_native(&model);
+                return NULL;
+            }
+            conv_params->kernel = av_malloc(kernel_size * sizeof(float));
+            conv_params->biases = av_malloc(conv_params->output_num * sizeof(float));
+            if (!conv_params->kernel || !conv_params->biases){
+                avio_closep(&model_file_context);
+                ff_dnn_free_model_native(&model);
+                return NULL;
+            }
+            for (i = 0; i < kernel_size; ++i){
+                conv_params->kernel[i] = av_int2float(avio_rl32(model_file_context));
+            }
+            for (i = 0; i < conv_params->output_num; ++i){
+                conv_params->biases[i] = av_int2float(avio_rl32(model_file_context));
+            }
+            network->layers[layer].type = CONV;
+            network->layers[layer].params = conv_params;
+            break;
+        case DEPTH_TO_SPACE:
+            depth_to_space_params = av_malloc(sizeof(DepthToSpaceParams));
+            if (!depth_to_space_params){
+                avio_closep(&model_file_context);
+                ff_dnn_free_model_native(&model);
+                return NULL;
+            }
+            depth_to_space_params->block_size = (int32_t)avio_rl32(model_file_context);
+            dnn_size += 4;
+            network->layers[layer].type = DEPTH_TO_SPACE;
+            network->layers[layer].params = depth_to_space_params;
+            break;
+        default:
+            avio_closep(&model_file_context);
+            ff_dnn_free_model_native(&model);
+            return NULL;
+        }
+    }
+
+    avio_closep(&model_file_context);
+
+    if (dnn_size != file_size){
+        ff_dnn_free_model_native(&model);
+        return NULL;
+    }
+
+    model->set_input_output = &set_input_output_native;
+
+    return model;
+}
+
+#define CLAMP_TO_EDGE(x, w) ((x) < 0 ? 0 : ((x) >= (w) ? (w - 1) : (x)))
+
+static void convolve(const float *input, float *output, const ConvolutionalParams *conv_params, int width, int height)
+{
+    int radius = conv_params->kernel_size >> 1;
+    int src_linesize = width * conv_params->input_num;
+    int filter_linesize = conv_params->kernel_size * conv_params->input_num;
+    int filter_size = conv_params->kernel_size * filter_linesize;
+    int pad_size = (conv_params->padding_method == VALID) ? (conv_params->kernel_size - 1) / 2 * conv_params->dilation : 0;
+
+    for (int y = pad_size; y < height - pad_size; ++y) {
+        for (int x = pad_size; x < width - pad_size; ++x) {
+            for (int n_filter = 0; n_filter < conv_params->output_num; ++n_filter) {
+                output[n_filter] = conv_params->biases[n_filter];
+
+                for (int ch = 0; ch < conv_params->input_num; ++ch) {
+                    for (int kernel_y = 0; kernel_y < conv_params->kernel_size; ++kernel_y) {
+                        for (int kernel_x = 0; kernel_x < conv_params->kernel_size; ++kernel_x) {
+                            float input_pel;
+                            if (conv_params->padding_method == SAME_CLAMP_TO_EDGE) {
+                                int y_pos = CLAMP_TO_EDGE(y + (kernel_y - radius) * conv_params->dilation, height);
+                                int x_pos = CLAMP_TO_EDGE(x + (kernel_x - radius) * conv_params->dilation, width);
+                                input_pel = input[y_pos * src_linesize + x_pos * conv_params->input_num + ch];
+                            } else {
+                                int y_pos = y + (kernel_y - radius) * conv_params->dilation;
+                                int x_pos = x + (kernel_x - radius) * conv_params->dilation;
+                                input_pel = (x_pos < 0 || x_pos >= width || y_pos < 0 || y_pos >= height) ? 0.0 :
+                                                   input[y_pos * src_linesize + x_pos * conv_params->input_num + ch];
+                            }
+
+
+                            output[n_filter] += input_pel * conv_params->kernel[n_filter * filter_size + kernel_y * filter_linesize +
+                                                                                kernel_x * conv_params->input_num + ch];
+                        }
+                    }
+                }
+                switch (conv_params->activation){
+                case RELU:
+                    output[n_filter] = FFMAX(output[n_filter], 0.0);
+                    break;
+                case TANH:
+                    output[n_filter] = 2.0f  / (1.0f + exp(-2.0f * output[n_filter])) - 1.0f;
+                    break;
+                case SIGMOID:
+                    output[n_filter] = 1.0f / (1.0f + exp(-output[n_filter]));
+                    break;
+                case NONE:
+                    break;
+                case LEAKY_RELU:
+                    output[n_filter] = FFMAX(output[n_filter], 0.0) + 0.2 * FFMIN(output[n_filter], 0.0);
+                }
+            }
+            output += conv_params->output_num;
+        }
+    }
+}
+
+static void depth_to_space(const float *input, float *output, int block_size, int width, int height, int channels)
+{
+    int y, x, by, bx, ch;
+    int new_channels = channels / (block_size * block_size);
+    int output_linesize = width * channels;
+    int by_linesize = output_linesize / block_size;
+    int x_linesize = new_channels * block_size;
+
+    for (y = 0; y < height; ++y){
+        for (x = 0; x < width; ++x){
+            for (by = 0; by < block_size; ++by){
+                for (bx = 0; bx < block_size; ++bx){
+                    for (ch = 0; ch < new_channels; ++ch){
+                        output[by * by_linesize + x * x_linesize + bx * new_channels + ch] = input[ch];
+                    }
+                    input += new_channels;
+                }
+            }
+        }
+        output += output_linesize;
+    }
+}
+
+DNNReturnType ff_dnn_execute_model_native(const DNNModel *model, DNNData *outputs, uint32_t nb_output)
+{
+    ConvolutionalNetwork *network = (ConvolutionalNetwork *)model->model;
+    int cur_width, cur_height, cur_channels;
+    int32_t layer;
+    InputParams *input_params;
+    ConvolutionalParams *conv_params;
+    DepthToSpaceParams *depth_to_space_params;
+
+    if (network->layers_num <= 0 || network->layers[0].type != INPUT || !network->layers[0].output){
+        return DNN_ERROR;
+    }
+    else{
+        input_params = (InputParams *)network->layers[0].params;
+        cur_width = input_params->width;
+        cur_height = input_params->height;
+        cur_channels = input_params->channels;
+    }
+
+    for (layer = 1; layer < network->layers_num; ++layer){
+        if (!network->layers[layer].output){
+            return DNN_ERROR;
+        }
+        switch (network->layers[layer].type){
+        case CONV:
+            conv_params = (ConvolutionalParams *)network->layers[layer].params;
+            convolve(network->layers[layer - 1].output, network->layers[layer].output, conv_params, cur_width, cur_height);
+            cur_channels = conv_params->output_num;
+            if (conv_params->padding_method == VALID) {
+                int pad_size = (conv_params->kernel_size - 1) * conv_params->dilation;
+                cur_height -= pad_size;
+                cur_width -= pad_size;
+            }
+            break;
+        case DEPTH_TO_SPACE:
+            depth_to_space_params = (DepthToSpaceParams *)network->layers[layer].params;
+            depth_to_space(network->layers[layer - 1].output, network->layers[layer].output,
+                           depth_to_space_params->block_size, cur_width, cur_height, cur_channels);
+            cur_height *= depth_to_space_params->block_size;
+            cur_width *= depth_to_space_params->block_size;
+            cur_channels /= depth_to_space_params->block_size * depth_to_space_params->block_size;
+            break;
+        case INPUT:
+            return DNN_ERROR;
+        }
+    }
+
+    // native mode does not support multiple outputs yet
+    if (nb_output > 1)
+        return DNN_ERROR;
+    outputs[0].data = network->layers[network->layers_num - 1].output;
+    outputs[0].height = cur_height;
+    outputs[0].width = cur_width;
+    outputs[0].channels = cur_channels;
+
+    return DNN_SUCCESS;
+}
+
+void ff_dnn_free_model_native(DNNModel **model)
+{
+    ConvolutionalNetwork *network;
+    ConvolutionalParams *conv_params;
+    int32_t layer;
+
+    if (*model)
+    {
+        network = (ConvolutionalNetwork *)(*model)->model;
+        for (layer = 0; layer < network->layers_num; ++layer){
+            av_freep(&network->layers[layer].output);
+            if (network->layers[layer].type == CONV){
+                conv_params = (ConvolutionalParams *)network->layers[layer].params;
+                av_freep(&conv_params->kernel);
+                av_freep(&conv_params->biases);
+            }
+            av_freep(&network->layers[layer].params);
+        }
+        av_freep(&network->layers);
+        av_freep(&network);
+        av_freep(model);
+    }
+}
diff --git a/libavfilter/dnn/dnn_backend_native.h b/libavfilter/dnn/dnn_backend_native.h
new file mode 100644
index 0000000000..8ef1855e45
--- /dev/null
+++ b/libavfilter/dnn/dnn_backend_native.h
@@ -0,0 +1,74 @@
+/*
+ * Copyright (c) 2018 Sergey Lavrushkin
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * DNN inference functions interface for native backend.
+ */
+
+
+#ifndef AVFILTER_DNN_DNN_BACKEND_NATIVE_H
+#define AVFILTER_DNN_DNN_BACKEND_NATIVE_H
+
+#include "../dnn_interface.h"
+#include "libavformat/avio.h"
+
+typedef enum {INPUT, CONV, DEPTH_TO_SPACE} DNNLayerType;
+
+typedef enum {RELU, TANH, SIGMOID, NONE, LEAKY_RELU} DNNActivationFunc;
+
+typedef enum {VALID, SAME, SAME_CLAMP_TO_EDGE} DNNConvPaddingParam;
+
+typedef struct Layer{
+    DNNLayerType type;
+    float *output;
+    void *params;
+} Layer;
+
+typedef struct ConvolutionalParams{
+    int32_t input_num, output_num, kernel_size;
+    DNNActivationFunc activation;
+    DNNConvPaddingParam padding_method;
+    int32_t dilation;
+    float *kernel;
+    float *biases;
+} ConvolutionalParams;
+
+typedef struct InputParams{
+    int height, width, channels;
+} InputParams;
+
+typedef struct DepthToSpaceParams{
+    int block_size;
+} DepthToSpaceParams;
+
+// Represents simple feed-forward convolutional network.
+typedef struct ConvolutionalNetwork{
+    Layer *layers;
+    int32_t layers_num;
+} ConvolutionalNetwork;
+
+DNNModel *ff_dnn_load_model_native(const char *model_filename);
+
+DNNReturnType ff_dnn_execute_model_native(const DNNModel *model, DNNData *outputs, uint32_t nb_output);
+
+void ff_dnn_free_model_native(DNNModel **model);
+
+#endif
diff --git a/libavfilter/dnn/dnn_backend_tf.c b/libavfilter/dnn/dnn_backend_tf.c
new file mode 100644
index 0000000000..ba959ae3a2
--- /dev/null
+++ b/libavfilter/dnn/dnn_backend_tf.c
@@ -0,0 +1,603 @@
+/*
+ * Copyright (c) 2018 Sergey Lavrushkin
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * DNN tensorflow backend implementation.
+ */
+
+#include "dnn_backend_tf.h"
+#include "dnn_backend_native.h"
+#include "libavformat/avio.h"
+#include "libavutil/avassert.h"
+
+#include <tensorflow/c/c_api.h>
+
+typedef struct TFModel{
+    TF_Graph *graph;
+    TF_Session *session;
+    TF_Status *status;
+    TF_Output input;
+    TF_Tensor *input_tensor;
+    TF_Output *outputs;
+    TF_Tensor **output_tensors;
+    uint32_t nb_output;
+} TFModel;
+
+static void free_buffer(void *data, size_t length)
+{
+    av_freep(&data);
+}
+
+static TF_Buffer *read_graph(const char *model_filename)
+{
+    TF_Buffer *graph_buf;
+    unsigned char *graph_data = NULL;
+    AVIOContext *model_file_context;
+    long size, bytes_read;
+
+    if (avio_open(&model_file_context, model_filename, AVIO_FLAG_READ) < 0){
+        return NULL;
+    }
+
+    size = avio_size(model_file_context);
+
+    graph_data = av_malloc(size);
+    if (!graph_data){
+        avio_closep(&model_file_context);
+        return NULL;
+    }
+    bytes_read = avio_read(model_file_context, graph_data, size);
+    avio_closep(&model_file_context);
+    if (bytes_read != size){
+        av_freep(&graph_data);
+        return NULL;
+    }
+
+    graph_buf = TF_NewBuffer();
+    graph_buf->data = (void *)graph_data;
+    graph_buf->length = size;
+    graph_buf->data_deallocator = free_buffer;
+
+    return graph_buf;
+}
+
+static TF_Tensor *allocate_input_tensor(const DNNInputData *input)
+{
+    TF_DataType dt;
+    size_t size;
+    int64_t input_dims[] = {1, input->height, input->width, input->channels};
+    switch (input->dt) {
+    case DNN_FLOAT:
+        dt = TF_FLOAT;
+        size = sizeof(float);
+        break;
+    case DNN_UINT8:
+        dt = TF_UINT8;
+        size = sizeof(char);
+        break;
+    default:
+        av_assert0(!"should not reach here");
+    }
+
+    return TF_AllocateTensor(dt, input_dims, 4,
+                             input_dims[1] * input_dims[2] * input_dims[3] * size);
+}
+
+static DNNReturnType set_input_output_tf(void *model, DNNInputData *input, const char *input_name, const char **output_names, uint32_t nb_output)
+{
+    TFModel *tf_model = (TFModel *)model;
+    TF_SessionOptions *sess_opts;
+    const TF_Operation *init_op = TF_GraphOperationByName(tf_model->graph, "init");
+
+    // Input operation
+    tf_model->input.oper = TF_GraphOperationByName(tf_model->graph, input_name);
+    if (!tf_model->input.oper){
+        return DNN_ERROR;
+    }
+    tf_model->input.index = 0;
+    if (tf_model->input_tensor){
+        TF_DeleteTensor(tf_model->input_tensor);
+    }
+    tf_model->input_tensor = allocate_input_tensor(input);
+    if (!tf_model->input_tensor){
+        return DNN_ERROR;
+    }
+    input->data = (float *)TF_TensorData(tf_model->input_tensor);
+
+    // Output operation
+    if (nb_output == 0)
+        return DNN_ERROR;
+
+    av_freep(&tf_model->outputs);
+    tf_model->outputs = av_malloc_array(nb_output, sizeof(*tf_model->outputs));
+    if (!tf_model->outputs)
+        return DNN_ERROR;
+    for (int i = 0; i < nb_output; ++i) {
+        tf_model->outputs[i].oper = TF_GraphOperationByName(tf_model->graph, output_names[i]);
+        if (!tf_model->outputs[i].oper){
+            av_freep(&tf_model->outputs);
+            return DNN_ERROR;
+        }
+        tf_model->outputs[i].index = 0;
+    }
+
+    if (tf_model->output_tensors) {
+        for (uint32_t i = 0; i < tf_model->nb_output; ++i) {
+            if (tf_model->output_tensors[i]) {
+                TF_DeleteTensor(tf_model->output_tensors[i]);
+                tf_model->output_tensors[i] = NULL;
+            }
+        }
+    }
+    av_freep(&tf_model->output_tensors);
+    tf_model->output_tensors = av_mallocz_array(nb_output, sizeof(*tf_model->output_tensors));
+    if (!tf_model->output_tensors) {
+        av_freep(&tf_model->outputs);
+        return DNN_ERROR;
+    }
+
+    tf_model->nb_output = nb_output;
+
+    if (tf_model->session){
+        TF_CloseSession(tf_model->session, tf_model->status);
+        TF_DeleteSession(tf_model->session, tf_model->status);
+    }
+
+    sess_opts = TF_NewSessionOptions();
+    tf_model->session = TF_NewSession(tf_model->graph, sess_opts, tf_model->status);
+    TF_DeleteSessionOptions(sess_opts);
+    if (TF_GetCode(tf_model->status) != TF_OK)
+    {
+        return DNN_ERROR;
+    }
+
+    // Run initialization operation with name "init" if it is present in graph
+    if (init_op){
+        TF_SessionRun(tf_model->session, NULL,
+                      NULL, NULL, 0,
+                      NULL, NULL, 0,
+                      &init_op, 1, NULL, tf_model->status);
+        if (TF_GetCode(tf_model->status) != TF_OK)
+        {
+            return DNN_ERROR;
+        }
+    }
+
+    return DNN_SUCCESS;
+}
+
+static DNNReturnType load_tf_model(TFModel *tf_model, const char *model_filename)
+{
+    TF_Buffer *graph_def;
+    TF_ImportGraphDefOptions *graph_opts;
+
+    graph_def = read_graph(model_filename);
+    if (!graph_def){
+        return DNN_ERROR;
+    }
+    tf_model->graph = TF_NewGraph();
+    tf_model->status = TF_NewStatus();
+    graph_opts = TF_NewImportGraphDefOptions();
+    TF_GraphImportGraphDef(tf_model->graph, graph_def, graph_opts, tf_model->status);
+    TF_DeleteImportGraphDefOptions(graph_opts);
+    TF_DeleteBuffer(graph_def);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        TF_DeleteGraph(tf_model->graph);
+        TF_DeleteStatus(tf_model->status);
+        return DNN_ERROR;
+    }
+
+    return DNN_SUCCESS;
+}
+
+#define NAME_BUFFER_SIZE 256
+
+static DNNReturnType add_conv_layer(TFModel *tf_model, TF_Operation *transpose_op, TF_Operation **cur_op,
+                                    ConvolutionalParams* params, const int layer)
+{
+    TF_Operation *op;
+    TF_OperationDescription *op_desc;
+    TF_Output input;
+    int64_t strides[] = {1, 1, 1, 1};
+    TF_Tensor *tensor;
+    int64_t dims[4];
+    int dims_len;
+    char name_buffer[NAME_BUFFER_SIZE];
+    int32_t size;
+
+    size = params->input_num * params->output_num * params->kernel_size * params->kernel_size;
+    input.index = 0;
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "conv_kernel%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "Const", name_buffer);
+    TF_SetAttrType(op_desc, "dtype", TF_FLOAT);
+    dims[0] = params->output_num;
+    dims[1] = params->kernel_size;
+    dims[2] = params->kernel_size;
+    dims[3] = params->input_num;
+    dims_len = 4;
+    tensor = TF_AllocateTensor(TF_FLOAT, dims, dims_len, size * sizeof(float));
+    memcpy(TF_TensorData(tensor), params->kernel, size * sizeof(float));
+    TF_SetAttrTensor(op_desc, "value", tensor, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+    op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "transpose%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "Transpose", name_buffer);
+    input.oper = op;
+    TF_AddInput(op_desc, input);
+    input.oper = transpose_op;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    TF_SetAttrType(op_desc, "Tperm", TF_INT32);
+    op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "conv2d%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "Conv2D", name_buffer);
+    input.oper = *cur_op;
+    TF_AddInput(op_desc, input);
+    input.oper = op;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    TF_SetAttrIntList(op_desc, "strides", strides, 4);
+    TF_SetAttrString(op_desc, "padding", "VALID", 5);
+    *cur_op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "conv_biases%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "Const", name_buffer);
+    TF_SetAttrType(op_desc, "dtype", TF_FLOAT);
+    dims[0] = params->output_num;
+    dims_len = 1;
+    tensor = TF_AllocateTensor(TF_FLOAT, dims, dims_len, params->output_num * sizeof(float));
+    memcpy(TF_TensorData(tensor), params->biases, params->output_num * sizeof(float));
+    TF_SetAttrTensor(op_desc, "value", tensor, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+    op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "bias_add%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "BiasAdd", name_buffer);
+    input.oper = *cur_op;
+    TF_AddInput(op_desc, input);
+    input.oper = op;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    *cur_op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "activation%d", layer);
+    switch (params->activation){
+    case RELU:
+        op_desc = TF_NewOperation(tf_model->graph, "Relu", name_buffer);
+        break;
+    case TANH:
+        op_desc = TF_NewOperation(tf_model->graph, "Tanh", name_buffer);
+        break;
+    case SIGMOID:
+        op_desc = TF_NewOperation(tf_model->graph, "Sigmoid", name_buffer);
+        break;
+    default:
+        return DNN_ERROR;
+    }
+    input.oper = *cur_op;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    *cur_op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    return DNN_SUCCESS;
+}
+
+static DNNReturnType add_depth_to_space_layer(TFModel *tf_model, TF_Operation **cur_op,
+                                              DepthToSpaceParams *params, const int layer)
+{
+    TF_OperationDescription *op_desc;
+    TF_Output input;
+    char name_buffer[NAME_BUFFER_SIZE];
+
+    snprintf(name_buffer, NAME_BUFFER_SIZE, "depth_to_space%d", layer);
+    op_desc = TF_NewOperation(tf_model->graph, "DepthToSpace", name_buffer);
+    input.oper = *cur_op;
+    input.index = 0;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    TF_SetAttrInt(op_desc, "block_size", params->block_size);
+    *cur_op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    return DNN_SUCCESS;
+}
+
+static int calculate_pad(const ConvolutionalNetwork *conv_network)
+{
+    ConvolutionalParams *params;
+    int32_t layer;
+    int pad = 0;
+
+    for (layer = 0; layer < conv_network->layers_num; ++layer){
+        if (conv_network->layers[layer].type == CONV){
+            params = (ConvolutionalParams *)conv_network->layers[layer].params;
+            pad += params->kernel_size >> 1;
+        }
+    }
+
+    return pad;
+}
+
+static DNNReturnType add_pad_op(TFModel *tf_model, TF_Operation **cur_op, const int32_t pad)
+{
+    TF_Operation *op;
+    TF_Tensor *tensor;
+    TF_OperationDescription *op_desc;
+    TF_Output input;
+    int32_t *pads;
+    int64_t pads_shape[] = {4, 2};
+
+    input.index = 0;
+
+    op_desc = TF_NewOperation(tf_model->graph, "Const", "pads");
+    TF_SetAttrType(op_desc, "dtype", TF_INT32);
+    tensor = TF_AllocateTensor(TF_INT32, pads_shape, 2, 4 * 2 * sizeof(int32_t));
+    pads = (int32_t *)TF_TensorData(tensor);
+    pads[0] = 0;   pads[1] = 0;
+    pads[2] = pad; pads[3] = pad;
+    pads[4] = pad; pads[5] = pad;
+    pads[6] = 0;   pads[7] = 0;
+    TF_SetAttrTensor(op_desc, "value", tensor, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+    op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    op_desc = TF_NewOperation(tf_model->graph, "MirrorPad", "mirror_pad");
+    input.oper = *cur_op;
+    TF_AddInput(op_desc, input);
+    input.oper = op;
+    TF_AddInput(op_desc, input);
+    TF_SetAttrType(op_desc, "T", TF_FLOAT);
+    TF_SetAttrType(op_desc, "Tpaddings", TF_INT32);
+    TF_SetAttrString(op_desc, "mode", "SYMMETRIC", 9);
+    *cur_op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    return DNN_SUCCESS;
+}
+
+static DNNReturnType load_native_model(TFModel *tf_model, const char *model_filename)
+{
+    int32_t layer;
+    TF_OperationDescription *op_desc;
+    TF_Operation *op;
+    TF_Operation *transpose_op;
+    TF_Tensor *tensor;
+    TF_Output input;
+    int32_t *transpose_perm;
+    int64_t transpose_perm_shape[] = {4};
+    int64_t input_shape[] = {1, -1, -1, -1};
+    int32_t pad;
+    DNNReturnType layer_add_res;
+    DNNModel *native_model = NULL;
+    ConvolutionalNetwork *conv_network;
+
+    native_model = ff_dnn_load_model_native(model_filename);
+    if (!native_model){
+        return DNN_ERROR;
+    }
+
+    conv_network = (ConvolutionalNetwork *)native_model->model;
+    pad = calculate_pad(conv_network);
+    tf_model->graph = TF_NewGraph();
+    tf_model->status = TF_NewStatus();
+
+#define CLEANUP_ON_ERROR(tf_model) \
+    { \
+        TF_DeleteGraph(tf_model->graph); \
+        TF_DeleteStatus(tf_model->status); \
+        return DNN_ERROR; \
+    }
+
+    op_desc = TF_NewOperation(tf_model->graph, "Placeholder", "x");
+    TF_SetAttrType(op_desc, "dtype", TF_FLOAT);
+    TF_SetAttrShape(op_desc, "shape", input_shape, 4);
+    op = TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        CLEANUP_ON_ERROR(tf_model);
+    }
+
+    if (add_pad_op(tf_model, &op, pad) != DNN_SUCCESS){
+        CLEANUP_ON_ERROR(tf_model);
+    }
+
+    op_desc = TF_NewOperation(tf_model->graph, "Const", "transpose_perm");
+    TF_SetAttrType(op_desc, "dtype", TF_INT32);
+    tensor = TF_AllocateTensor(TF_INT32, transpose_perm_shape, 1, 4 * sizeof(int32_t));
+    transpose_perm = (int32_t *)TF_TensorData(tensor);
+    transpose_perm[0] = 1;
+    transpose_perm[1] = 2;
+    transpose_perm[2] = 3;
+    transpose_perm[3] = 0;
+    TF_SetAttrTensor(op_desc, "value", tensor, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        CLEANUP_ON_ERROR(tf_model);
+    }
+    transpose_op = TF_FinishOperation(op_desc, tf_model->status);
+
+    for (layer = 0; layer < conv_network->layers_num; ++layer){
+        switch (conv_network->layers[layer].type){
+        case INPUT:
+            layer_add_res = DNN_SUCCESS;
+            break;
+        case CONV:
+            layer_add_res = add_conv_layer(tf_model, transpose_op, &op,
+                                           (ConvolutionalParams *)conv_network->layers[layer].params, layer);
+            break;
+        case DEPTH_TO_SPACE:
+            layer_add_res = add_depth_to_space_layer(tf_model, &op,
+                                                     (DepthToSpaceParams *)conv_network->layers[layer].params, layer);
+            break;
+        default:
+            CLEANUP_ON_ERROR(tf_model);
+        }
+
+        if (layer_add_res != DNN_SUCCESS){
+            CLEANUP_ON_ERROR(tf_model);
+        }
+    }
+
+    op_desc = TF_NewOperation(tf_model->graph, "Identity", "y");
+    input.oper = op;
+    TF_AddInput(op_desc, input);
+    TF_FinishOperation(op_desc, tf_model->status);
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        CLEANUP_ON_ERROR(tf_model);
+    }
+
+    ff_dnn_free_model_native(&native_model);
+
+    return DNN_SUCCESS;
+}
+
+DNNModel *ff_dnn_load_model_tf(const char *model_filename)
+{
+    DNNModel *model = NULL;
+    TFModel *tf_model = NULL;
+
+    model = av_malloc(sizeof(DNNModel));
+    if (!model){
+        return NULL;
+    }
+
+    tf_model = av_mallocz(sizeof(TFModel));
+    if (!tf_model){
+        av_freep(&model);
+        return NULL;
+    }
+
+    if (load_tf_model(tf_model, model_filename) != DNN_SUCCESS){
+        if (load_native_model(tf_model, model_filename) != DNN_SUCCESS){
+            av_freep(&tf_model);
+            av_freep(&model);
+
+            return NULL;
+        }
+    }
+
+    model->model = (void *)tf_model;
+    model->set_input_output = &set_input_output_tf;
+
+    return model;
+}
+
+
+
+DNNReturnType ff_dnn_execute_model_tf(const DNNModel *model, DNNData *outputs, uint32_t nb_output)
+{
+    TFModel *tf_model = (TFModel *)model->model;
+    uint32_t nb = FFMIN(nb_output, tf_model->nb_output);
+    if (nb == 0)
+        return DNN_ERROR;
+
+    av_assert0(tf_model->output_tensors);
+    for (uint32_t i = 0; i < tf_model->nb_output; ++i) {
+        if (tf_model->output_tensors[i]) {
+            TF_DeleteTensor(tf_model->output_tensors[i]);
+            tf_model->output_tensors[i] = NULL;
+        }
+    }
+
+    TF_SessionRun(tf_model->session, NULL,
+                  &tf_model->input, &tf_model->input_tensor, 1,
+                  tf_model->outputs, tf_model->output_tensors, nb,
+                  NULL, 0, NULL, tf_model->status);
+
+    if (TF_GetCode(tf_model->status) != TF_OK){
+        return DNN_ERROR;
+    }
+
+    for (uint32_t i = 0; i < nb; ++i) {
+        outputs[i].height = TF_Dim(tf_model->output_tensors[i], 1);
+        outputs[i].width = TF_Dim(tf_model->output_tensors[i], 2);
+        outputs[i].channels = TF_Dim(tf_model->output_tensors[i], 3);
+        outputs[i].data = TF_TensorData(tf_model->output_tensors[i]);
+    }
+
+    return DNN_SUCCESS;
+}
+
+void ff_dnn_free_model_tf(DNNModel **model)
+{
+    TFModel *tf_model;
+
+    if (*model){
+        tf_model = (TFModel *)(*model)->model;
+        if (tf_model->graph){
+            TF_DeleteGraph(tf_model->graph);
+        }
+        if (tf_model->session){
+            TF_CloseSession(tf_model->session, tf_model->status);
+            TF_DeleteSession(tf_model->session, tf_model->status);
+        }
+        if (tf_model->status){
+            TF_DeleteStatus(tf_model->status);
+        }
+        if (tf_model->input_tensor){
+            TF_DeleteTensor(tf_model->input_tensor);
+        }
+        if (tf_model->output_tensors) {
+            for (uint32_t i = 0; i < tf_model->nb_output; ++i) {
+                if (tf_model->output_tensors[i]) {
+                    TF_DeleteTensor(tf_model->output_tensors[i]);
+                    tf_model->output_tensors[i] = NULL;
+                }
+            }
+        }
+        av_freep(&tf_model->outputs);
+        av_freep(&tf_model->output_tensors);
+        av_freep(&tf_model);
+        av_freep(model);
+    }
+}
diff --git a/libavfilter/dnn/dnn_backend_tf.h b/libavfilter/dnn/dnn_backend_tf.h
new file mode 100644
index 0000000000..3e4508912e
--- /dev/null
+++ b/libavfilter/dnn/dnn_backend_tf.h
@@ -0,0 +1,38 @@
+/*
+ * Copyright (c) 2018 Sergey Lavrushkin
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * DNN inference functions interface for TensorFlow backend.
+ */
+
+
+#ifndef AVFILTER_DNN_DNN_BACKEND_TF_H
+#define AVFILTER_DNN_DNN_BACKEND_TF_H
+
+#include "../dnn_interface.h"
+
+DNNModel *ff_dnn_load_model_tf(const char *model_filename);
+
+DNNReturnType ff_dnn_execute_model_tf(const DNNModel *model, DNNData *outputs, uint32_t nb_output);
+
+void ff_dnn_free_model_tf(DNNModel **model);
+
+#endif
diff --git a/libavfilter/dnn/dnn_interface.c b/libavfilter/dnn/dnn_interface.c
new file mode 100644
index 0000000000..62da55f43e
--- /dev/null
+++ b/libavfilter/dnn/dnn_interface.c
@@ -0,0 +1,63 @@
+/*
+ * Copyright (c) 2018 Sergey Lavrushkin
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * Implements DNN module initialization with specified backend.
+ */
+
+#include "../dnn_interface.h"
+#include "dnn_backend_native.h"
+#include "dnn_backend_tf.h"
+#include "libavutil/mem.h"
+
+DNNModule *ff_get_dnn_module(DNNBackendType backend_type)
+{
+    DNNModule *dnn_module;
+
+    dnn_module = av_malloc(sizeof(DNNModule));
+    if(!dnn_module){
+        return NULL;
+    }
+
+    switch(backend_type){
+    case DNN_NATIVE:
+        dnn_module->load_model = &ff_dnn_load_model_native;
+        dnn_module->execute_model = &ff_dnn_execute_model_native;
+        dnn_module->free_model = &ff_dnn_free_model_native;
+        break;
+    case DNN_TF:
+    #if (CONFIG_LIBTENSORFLOW == 1)
+        dnn_module->load_model = &ff_dnn_load_model_tf;
+        dnn_module->execute_model = &ff_dnn_execute_model_tf;
+        dnn_module->free_model = &ff_dnn_free_model_tf;
+    #else
+        av_freep(&dnn_module);
+        return NULL;
+    #endif
+        break;
+    default:
+        av_log(NULL, AV_LOG_ERROR, "Module backend_type is not native or tensorflow\n");
+        av_freep(&dnn_module);
+        return NULL;
+    }
+
+    return dnn_module;
+}