diff mbox series

[FFmpeg-devel,V6,2/3] lavfi/dnn: Modified DNN native backend related tools and docs.

Message ID 20230306135548.23001-2-ting.fu@intel.com
State New
Headers show
Series [FFmpeg-devel,V6,1/3] lavfi/dnn: Mark native backend as unsupported | expand

Checks

Context Check Description
yinshiyou/make_loongarch64 success Make finished
yinshiyou/make_fate_loongarch64 success Make fate finished
andriy/make_x86 success Make finished
andriy/make_fate_x86 success Make fate finished

Commit Message

Ting Fu March 6, 2023, 1:55 p.m. UTC
Deleted the native backend related files in 'tools' dir. Modify its'
docs and codes mentioned in such docs.

Signed-off-by: Ting Fu <ting.fu@intel.com>
---
 doc/filters.texi                        |  39 +-
 libavfilter/vf_derain.c                 |   2 +-
 libavfilter/vf_dnn_processing.c         |   2 +-
 libavfilter/vf_sr.c                     |   2 +-
 tools/python/convert.py                 |  56 ---
 tools/python/convert_from_tensorflow.py | 607 ------------------------
 tools/python/convert_header.py          |  26 -
 7 files changed, 7 insertions(+), 727 deletions(-)
 delete mode 100644 tools/python/convert.py
 delete mode 100644 tools/python/convert_from_tensorflow.py
 delete mode 100644 tools/python/convert_header.py

Comments

Guo, Yejun April 27, 2023, 3:23 a.m. UTC | #1
> -----Original Message-----
> From: ffmpeg-devel <ffmpeg-devel-bounces@ffmpeg.org> On Behalf Of
> Ting Fu
> Sent: Monday, March 6, 2023 9:56 PM
> To: ffmpeg-devel@ffmpeg.org
> Subject: [FFmpeg-devel] [PATCH V6 2/3] lavfi/dnn: Modified DNN native
> backend related tools and docs.
> 
> Deleted the native backend related files in 'tools' dir. Modify its'
> docs and codes mentioned in such docs.

We'll remove native backend, so change the default backend in filters, 
and also remove the python scripts which generate native model file.
diff mbox series

Patch

diff --git a/doc/filters.texi b/doc/filters.texi
index 7a7b2ba4e7..726d2fd7e2 100644
--- a/doc/filters.texi
+++ b/doc/filters.texi
@@ -11305,9 +11305,6 @@  See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_
 Training as well as model generation scripts are provided in
 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
 
-Native model files (.model) can be generated from TensorFlow model
-files (.pb) by using tools/python/convert.py
-
 The filter accepts the following options:
 
 @table @option
@@ -11328,21 +11325,16 @@  Specify which DNN backend to use for model loading and execution. This option ac
 the following values:
 
 @table @samp
-@item native
-Native implementation of DNN loading and execution.
-
 @item tensorflow
 TensorFlow backend. To enable this backend you
 need to install the TensorFlow for C library (see
 @url{https://www.tensorflow.org/install/lang_c}) and configure FFmpeg with
 @code{--enable-libtensorflow}
 @end table
-Default value is @samp{native}.
 
 @item model
 Set path to model file specifying network architecture and its parameters.
-Note that different backends use different file formats. TensorFlow and native
-backend can load files for only its format.
+Note that different backends use different file formats. TensorFlow can load files for only its format.
 @end table
 
 To get full functionality (such as async execution), please use the @ref{dnn_processing} filter.
@@ -11666,9 +11658,6 @@  Specify which DNN backend to use for model loading and execution. This option ac
 the following values:
 
 @table @samp
-@item native
-Native implementation of DNN loading and execution.
-
 @item tensorflow
 TensorFlow backend. To enable this backend you
 need to install the TensorFlow for C library (see
@@ -11684,14 +11673,9 @@  be needed if the header files and libraries are not installed into system path)
 
 @end table
 
-Default value is @samp{native}.
-
 @item model
 Set path to model file specifying network architecture and its parameters.
-Note that different backends use different file formats. TensorFlow, OpenVINO and native
-backend can load files for only its format.
-
-Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
+Note that different backends use different file formats. TensorFlow, OpenVINO backend can load files for only its format.
 
 @item input
 Set the input name of the dnn network.
@@ -11717,12 +11701,6 @@  Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
 @end example
 
-@item
-Halve the pixel value of the frame with format gray32f:
-@example
-ffmpeg -i input.jpg -vf format=grayf32,dnn_processing=model=halve_gray_float.model:input=dnn_in:output=dnn_out:dnn_backend=native -y out.native.png
-@end example
-
 @item
 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
 @example
@@ -21750,9 +21728,6 @@  Training scripts as well as scripts for model file (.pb) saving can be found at
 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
 
-Native model files (.model) can be generated from TensorFlow model
-files (.pb) by using tools/python/convert.py
-
 The filter accepts the following options:
 
 @table @option
@@ -21761,9 +21736,6 @@  Specify which DNN backend to use for model loading and execution. This option ac
 the following values:
 
 @table @samp
-@item native
-Native implementation of DNN loading and execution.
-
 @item tensorflow
 TensorFlow backend. To enable this backend you
 need to install the TensorFlow for C library (see
@@ -21771,13 +21743,10 @@  need to install the TensorFlow for C library (see
 @code{--enable-libtensorflow}
 @end table
 
-Default value is @samp{native}.
-
 @item model
 Set path to model file specifying network architecture and its parameters.
-Note that different backends use different file formats. TensorFlow backend
-can load files for both formats, while native backend can load files for only
-its format.
+Note that different backends use different file formats. TensorFlow, OpenVINO backend
+can load files for only its format.
 
 @item scale_factor
 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
diff --git a/libavfilter/vf_derain.c b/libavfilter/vf_derain.c
index 86e9eb8752..7e84cd65a3 100644
--- a/libavfilter/vf_derain.c
+++ b/libavfilter/vf_derain.c
@@ -43,7 +43,7 @@  static const AVOption derain_options[] = {
     { "filter_type", "filter type(derain/dehaze)",  OFFSET(filter_type),    AV_OPT_TYPE_INT,    { .i64 = 0 },    0, 1, FLAGS, "type" },
     { "derain",      "derain filter flag",          0,                      AV_OPT_TYPE_CONST,  { .i64 = 0 },    0, 0, FLAGS, "type" },
     { "dehaze",      "dehaze filter flag",          0,                      AV_OPT_TYPE_CONST,  { .i64 = 1 },    0, 0, FLAGS, "type" },
-    { "dnn_backend", "DNN backend",                 OFFSET(dnnctx.backend_type),   AV_OPT_TYPE_INT,    { .i64 = 0 },    0, 1, FLAGS, "backend" },
+    { "dnn_backend", "DNN backend",                 OFFSET(dnnctx.backend_type),   AV_OPT_TYPE_INT,    { .i64 = 1 },    0, 1, FLAGS, "backend" },
     { "native",      "native backend flag",         0,                      AV_OPT_TYPE_CONST,  { .i64 = 0 },    0, 0, FLAGS, "backend" },
 #if (CONFIG_LIBTENSORFLOW == 1)
     { "tensorflow",  "tensorflow backend flag",     0,                      AV_OPT_TYPE_CONST,  { .i64 = 1 },    0, 0, FLAGS, "backend" },
diff --git a/libavfilter/vf_dnn_processing.c b/libavfilter/vf_dnn_processing.c
index 4462915073..28937346b5 100644
--- a/libavfilter/vf_dnn_processing.c
+++ b/libavfilter/vf_dnn_processing.c
@@ -45,7 +45,7 @@  typedef struct DnnProcessingContext {
 #define OFFSET(x) offsetof(DnnProcessingContext, dnnctx.x)
 #define FLAGS AV_OPT_FLAG_FILTERING_PARAM | AV_OPT_FLAG_VIDEO_PARAM
 static const AVOption dnn_processing_options[] = {
-    { "dnn_backend", "DNN backend",                OFFSET(backend_type),     AV_OPT_TYPE_INT,       { .i64 = 0 },    INT_MIN, INT_MAX, FLAGS, "backend" },
+    { "dnn_backend", "DNN backend",                OFFSET(backend_type),     AV_OPT_TYPE_INT,       { .i64 = 2 },    INT_MIN, INT_MAX, FLAGS, "backend" },
     { "native",      "native backend flag",        0,                        AV_OPT_TYPE_CONST,     { .i64 = 0 },    0, 0, FLAGS, "backend" },
 #if (CONFIG_LIBTENSORFLOW == 1)
     { "tensorflow",  "tensorflow backend flag",    0,                        AV_OPT_TYPE_CONST,     { .i64 = 1 },    0, 0, FLAGS, "backend" },
diff --git a/libavfilter/vf_sr.c b/libavfilter/vf_sr.c
index cb24c096ce..e9fe746bae 100644
--- a/libavfilter/vf_sr.c
+++ b/libavfilter/vf_sr.c
@@ -46,7 +46,7 @@  typedef struct SRContext {
 #define OFFSET(x) offsetof(SRContext, x)
 #define FLAGS AV_OPT_FLAG_FILTERING_PARAM | AV_OPT_FLAG_VIDEO_PARAM
 static const AVOption sr_options[] = {
-    { "dnn_backend", "DNN backend used for model execution", OFFSET(dnnctx.backend_type), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, FLAGS, "backend" },
+    { "dnn_backend", "DNN backend used for model execution", OFFSET(dnnctx.backend_type), AV_OPT_TYPE_INT, { .i64 = 1 }, 0, 1, FLAGS, "backend" },
     { "native", "native backend flag", 0, AV_OPT_TYPE_CONST, { .i64 = 0 }, 0, 0, FLAGS, "backend" },
 #if (CONFIG_LIBTENSORFLOW == 1)
     { "tensorflow", "tensorflow backend flag", 0, AV_OPT_TYPE_CONST, { .i64 = 1 }, 0, 0, FLAGS, "backend" },
diff --git a/tools/python/convert.py b/tools/python/convert.py
deleted file mode 100644
index 64cf76b2d8..0000000000
--- a/tools/python/convert.py
+++ /dev/null
@@ -1,56 +0,0 @@ 
-# Copyright (c) 2019 Guo Yejun
-#
-# 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
-# ==============================================================================
-
-# verified with Python 3.5.2 on Ubuntu 16.04
-import argparse
-import os
-from convert_from_tensorflow import *
-
-def get_arguments():
-    parser = argparse.ArgumentParser(description='generate native mode model with weights from deep learning model')
-    parser.add_argument('--outdir', type=str, default='./', help='where to put generated files')
-    parser.add_argument('--infmt', type=str, default='tensorflow', help='format of the deep learning model')
-    parser.add_argument('infile', help='path to the deep learning model with weights')
-    parser.add_argument('--dump4tb', type=str, default='no', help='dump file for visualization in tensorboard')
-
-    return parser.parse_args()
-
-def main():
-    args = get_arguments()
-
-    if not os.path.isfile(args.infile):
-        print('the specified input file %s does not exist' % args.infile)
-        exit(1)
-
-    if not os.path.exists(args.outdir):
-        print('create output directory %s' % args.outdir)
-        os.mkdir(args.outdir)
-
-    basefile = os.path.split(args.infile)[1]
-    basefile = os.path.splitext(basefile)[0]
-    outfile = os.path.join(args.outdir, basefile) + '.model'
-    dump4tb = False
-    if args.dump4tb.lower() in ('yes', 'true', 't', 'y', '1'):
-        dump4tb = True
-
-    if args.infmt == 'tensorflow':
-        convert_from_tensorflow(args.infile, outfile, dump4tb)
-
-if __name__ == '__main__':
-    main()
diff --git a/tools/python/convert_from_tensorflow.py b/tools/python/convert_from_tensorflow.py
deleted file mode 100644
index 38e64c1c94..0000000000
--- a/tools/python/convert_from_tensorflow.py
+++ /dev/null
@@ -1,607 +0,0 @@ 
-# Copyright (c) 2019 Guo Yejun
-#
-# 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
-# ==============================================================================
-
-import tensorflow as tf
-import numpy as np
-import sys, struct
-import convert_header as header
-
-__all__ = ['convert_from_tensorflow']
-
-class Operand(object):
-    IOTYPE_INPUT = 1
-    IOTYPE_OUTPUT = 2
-    IOTYPE_INTERMEDIATE = IOTYPE_INPUT | IOTYPE_OUTPUT
-    DTYPE_FLOAT = 1
-    DTYPE_UINT8 = 4
-    index = 0
-    def __init__(self, name, dtype, dims):
-        self.name = name
-        self.dtype = dtype
-        self.dims = dims
-        self.iotype = 0
-        self.used_count = 0
-        self.index = Operand.index
-        Operand.index = Operand.index + 1
-        self.iotype2str = {Operand.IOTYPE_INPUT: 'in', Operand.IOTYPE_OUTPUT: 'out', Operand.IOTYPE_INTERMEDIATE: 'inout'}
-        self.dtype2str = {Operand.DTYPE_FLOAT: 'DT_FLOAT', Operand.DTYPE_UINT8: 'DT_UINT8'}
-
-    def add_iotype(self, iotype):
-        self.iotype = self.iotype | iotype
-        if iotype == Operand.IOTYPE_INPUT:
-            self.used_count = self.used_count + 1
-
-    def __str__(self):
-        return "{}: (name: {}, iotype: {}, dtype: {}, dims: {}, used_count: {})".format(self.index,
-                            self.name, self.iotype2str[self.iotype], self.dtype2str[self.dtype],
-                            self.dims, self.used_count)
-
-    def __lt__(self, other):
-        return self.index < other.index
-
-class TFConverter:
-    def __init__(self, graph_def, nodes, outfile, dump4tb):
-        self.graph_def = graph_def
-        self.nodes = nodes
-        self.outfile = outfile
-        self.dump4tb = dump4tb
-        self.layer_number = 0
-        self.output_names = []
-        self.name_node_dict = {}
-        self.edges = {}
-        self.conv_activations = {'Relu':0, 'Tanh':1, 'Sigmoid':2, 'None':3, 'LeakyRelu':4}
-        self.conv_paddings = {'VALID':0, 'SAME':1}
-        self.pool_paddings = {'VALID':0, 'SAME':1}
-        self.converted_nodes = set()
-        self.conv2d_scope_names = set()
-        self.conv2d_scopename_inputname_dict = {}
-        self.dense_scope_names = set()
-        self.dense_scopename_inputname_dict = {}
-        self.op2code = {'Conv2D':1, 'DepthToSpace':2, 'MirrorPad':3, 'Maximum':4,
-                        'MathBinary':5, 'MathUnary':6, 'AvgPool':7, 'MatMul':8}
-        self.mathbin2code = {'Sub':0, 'Add':1, 'Mul':2, 'RealDiv':3, 'Minimum':4, 'FloorMod':5}
-        self.mathun2code  = {'Abs':0, 'Sin':1, 'Cos':2, 'Tan':3, 'Asin':4,
-                'Acos':5, 'Atan':6, 'Sinh':7, 'Cosh':8, 'Tanh':9, 'Asinh':10,
-                'Acosh':11, 'Atanh':12, 'Ceil':13, 'Floor':14, 'Round':15,
-                'Exp':16}
-        self.mirrorpad_mode = {'CONSTANT':0, 'REFLECT':1, 'SYMMETRIC':2}
-        self.name_operand_dict = {}
-
-
-    def add_operand(self, name, type):
-        node = self.name_node_dict[name]
-        if name not in self.name_operand_dict:
-            dtype = node.attr['dtype'].type
-            if dtype == 0:
-                dtype = node.attr['T'].type
-            dims = [-1,-1,-1,-1]
-            if 'shape' in node.attr:
-                dims[0] = node.attr['shape'].shape.dim[0].size
-                dims[1] = node.attr['shape'].shape.dim[1].size
-                dims[2] = node.attr['shape'].shape.dim[2].size
-                dims[3] = node.attr['shape'].shape.dim[3].size
-            operand = Operand(name, dtype, dims)
-            self.name_operand_dict[name] = operand;
-        self.name_operand_dict[name].add_iotype(type)
-        return self.name_operand_dict[name].index
-
-
-    def dump_for_tensorboard(self):
-        graph = tf.get_default_graph()
-        tf.import_graph_def(self.graph_def, name="")
-        tf.summary.FileWriter('/tmp/graph', graph)
-        print('graph saved, run "tensorboard --logdir=/tmp/graph" to see it')
-
-
-    def get_conv2d_params(self, conv2d_scope_name):
-        knode = self.name_node_dict[conv2d_scope_name + '/kernel']
-        bnode = self.name_node_dict[conv2d_scope_name + '/bias']
-
-        if conv2d_scope_name + '/dilation_rate' in self.name_node_dict:
-            dnode = self.name_node_dict[conv2d_scope_name + '/dilation_rate']
-        else:
-            dnode = None
-
-        # the BiasAdd name is possible be changed into the output name,
-        # if activation is None, and BiasAdd.next is the last op which is Identity
-        if conv2d_scope_name + '/BiasAdd' in self.edges:
-            anode = self.edges[conv2d_scope_name + '/BiasAdd'][0]
-            if anode.op not in self.conv_activations:
-                anode = None
-        else:
-            anode = None
-        return knode, bnode, dnode, anode
-
-
-    def get_dense_params(self, dense_scope_name):
-        knode = self.name_node_dict[dense_scope_name + '/kernel']
-        bnode = self.name_node_dict.get(dense_scope_name + '/bias')
-        # the BiasAdd name is possible be changed into the output name,
-        # if activation is None, and BiasAdd.next is the last op which is Identity
-        anode = None
-        if bnode:
-            if dense_scope_name + '/BiasAdd' in self.edges:
-                anode = self.edges[dense_scope_name + '/BiasAdd'][0]
-                if anode.op not in self.conv_activations:
-                    anode = None
-        else:
-            anode = None
-        return knode, bnode, anode
-
-
-    def dump_complex_conv2d_to_file(self, node, f):
-        assert(node.op == 'Conv2D')
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-
-        scope_name = TFConverter.get_scope_name(node.name)
-        #knode for kernel, bnode for bias, dnode for dilation, anode for activation
-        knode, bnode, dnode, anode = self.get_conv2d_params(scope_name)
-
-        if dnode is not None:
-            dilation = struct.unpack('i', dnode.attr['value'].tensor.tensor_content[0:4])[0]
-        else:
-            dilation = 1
-
-        if anode is not None:
-            activation = anode.op
-        else:
-            activation = 'None'
-
-        padding = node.attr['padding'].s.decode("utf-8")
-        # conv2d with dilation > 1 generates tens of nodes, not easy to parse them, so use this tricky method.
-        if dilation > 1 and scope_name + '/stack' in self.name_node_dict:
-            if self.name_node_dict[scope_name + '/stack'].op == "Const":
-                padding = 'SAME'
-        padding = self.conv_paddings[padding]
-
-        ktensor = knode.attr['value'].tensor
-        filter_height = ktensor.tensor_shape.dim[0].size
-        filter_width = ktensor.tensor_shape.dim[1].size
-        in_channels = ktensor.tensor_shape.dim[2].size
-        out_channels = ktensor.tensor_shape.dim[3].size
-        kernel = np.frombuffer(ktensor.tensor_content, dtype=np.float32)
-        kernel = kernel.reshape(filter_height, filter_width, in_channels, out_channels)
-        kernel = np.transpose(kernel, [3, 0, 1, 2])
-
-        has_bias = 1
-        np.array([self.op2code[node.op], dilation, padding, self.conv_activations[activation], in_channels, out_channels, filter_height, has_bias], dtype=np.uint32).tofile(f)
-        kernel.tofile(f)
-
-        btensor = bnode.attr['value'].tensor
-        if btensor.tensor_shape.dim[0].size == 1:
-            bias = struct.pack("f", btensor.float_val[0])
-        else:
-            bias = btensor.tensor_content
-        f.write(bias)
-
-        input_name = self.conv2d_scopename_inputname_dict[scope_name]
-        input_operand_index = self.add_operand(input_name, Operand.IOTYPE_INPUT)
-
-        if anode is not None:
-            output_operand_index = self.add_operand(anode.name, Operand.IOTYPE_OUTPUT)
-        else:
-            output_operand_index = self.add_operand(self.edges[bnode.name][0].name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-    def dump_dense_to_file(self, node, f):
-        assert(node.op == 'MatMul')
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-
-        scope_name = TFConverter.get_scope_name(node.name)
-        #knode for kernel, bnode for bias, anode for activation
-        knode, bnode, anode = self.get_dense_params(scope_name.split('/')[0])
-
-        if bnode is not None:
-            has_bias = 1
-            btensor = bnode.attr['value'].tensor
-            if btensor.tensor_shape.dim[0].size == 1:
-                bias = struct.pack("f", btensor.float_val[0])
-            else:
-                bias = btensor.tensor_content
-        else:
-            has_bias = 0
-
-        if anode is not None:
-            activation = anode.op
-        else:
-            activation = 'None'
-
-        ktensor = knode.attr['value'].tensor
-        in_channels = ktensor.tensor_shape.dim[0].size
-        out_channels = ktensor.tensor_shape.dim[1].size
-        if in_channels * out_channels == 1:
-            kernel = np.float32(ktensor.float_val[0])
-        else:
-            kernel = np.frombuffer(ktensor.tensor_content, dtype=np.float32)
-        kernel = kernel.reshape(in_channels, out_channels)
-        kernel = np.transpose(kernel, [1, 0])
-
-        np.array([self.op2code[node.op], self.conv_activations[activation], in_channels, out_channels, has_bias], dtype=np.uint32).tofile(f)
-        kernel.tofile(f)
-        if has_bias:
-            f.write(bias)
-
-        input_name = self.dense_scopename_inputname_dict[scope_name.split('/')[0]]
-        input_operand_index = self.add_operand(input_name, Operand.IOTYPE_INPUT)
-
-        if anode is not None:
-            output_operand_index = self.add_operand(anode.name, Operand.IOTYPE_OUTPUT)
-        else:
-            if bnode is not None:
-                output_operand_index = self.add_operand(self.edges[bnode.name][0].name, Operand.IOTYPE_OUTPUT)
-            else:
-                output_operand_index = self.add_operand(self.edges[scope_name+'/concat_1'][0].name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_simple_conv2d_to_file(self, node, f):
-        assert(node.op == 'Conv2D')
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-
-        node0 = self.name_node_dict[node.input[0]]
-        node1 = self.name_node_dict[node.input[1]]
-        if node0.op == 'Const':
-            knode = node0
-            input_name = node.input[1]
-        else:
-            knode = node1
-            input_name = node.input[0]
-
-        ktensor = knode.attr['value'].tensor
-        filter_height = ktensor.tensor_shape.dim[0].size
-        filter_width = ktensor.tensor_shape.dim[1].size
-        in_channels = ktensor.tensor_shape.dim[2].size
-        out_channels = ktensor.tensor_shape.dim[3].size
-        if filter_height * filter_width * in_channels * out_channels == 1:
-            kernel = np.float32(ktensor.float_val[0])
-        else:
-            kernel = np.frombuffer(ktensor.tensor_content, dtype=np.float32)
-        kernel = kernel.reshape(filter_height, filter_width, in_channels, out_channels)
-        kernel = np.transpose(kernel, [3, 0, 1, 2])
-
-        has_bias = 0
-        dilation = 1
-        padding = node.attr['padding'].s.decode("utf-8")
-        np.array([self.op2code[node.op], dilation, self.conv_paddings[padding], self.conv_activations['None'],
-                  in_channels, out_channels, filter_height, has_bias], dtype=np.uint32).tofile(f)
-        kernel.tofile(f)
-
-        input_operand_index = self.add_operand(input_name, Operand.IOTYPE_INPUT)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_depth2space_to_file(self, node, f):
-        assert(node.op == 'DepthToSpace')
-        self.layer_number = self.layer_number + 1
-        block_size = node.attr['block_size'].i
-        np.array([self.op2code[node.op], block_size], dtype=np.uint32).tofile(f)
-        self.converted_nodes.add(node.name)
-        input_operand_index = self.add_operand(node.input[0], Operand.IOTYPE_INPUT)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_mirrorpad_to_file(self, node, f):
-        assert(node.op == 'MirrorPad')
-        self.layer_number = self.layer_number + 1
-        mode = node.attr['mode'].s
-        mode = self.mirrorpad_mode[mode.decode("utf-8")]
-        np.array([self.op2code[node.op], mode], dtype=np.uint32).tofile(f)
-        pnode = self.name_node_dict[node.input[1]]
-        self.converted_nodes.add(pnode.name)
-        paddings = pnode.attr['value'].tensor.tensor_content
-        f.write(paddings)
-        self.converted_nodes.add(node.name)
-        input_operand_index = self.add_operand(node.input[0], Operand.IOTYPE_INPUT)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_maximum_to_file(self, node, f):
-        assert(node.op == 'Maximum')
-        self.layer_number = self.layer_number + 1
-        ynode = self.name_node_dict[node.input[1]]
-        y = ynode.attr['value'].tensor.float_val[0]
-        np.array([self.op2code[node.op]], dtype=np.uint32).tofile(f)
-        np.array([y], dtype=np.float32).tofile(f)
-        self.converted_nodes.add(node.name)
-        input_operand_index = self.add_operand(node.input[0], Operand.IOTYPE_INPUT)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_mathbinary_to_file(self, node, f):
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-        i0_node = self.name_node_dict[node.input[0]]
-        i1_node = self.name_node_dict[node.input[1]]
-        np.array([self.op2code['MathBinary'], self.mathbin2code[node.op]], dtype=np.uint32).tofile(f)
-        if i0_node.op == 'Const':
-            scalar = i0_node.attr['value'].tensor.float_val[0]
-            np.array([1], dtype=np.uint32).tofile(f)            # broadcast: 1
-            np.array([scalar], dtype=np.float32).tofile(f)
-            np.array([0], dtype=np.uint32).tofile(f)            # broadcast: 0
-            input_operand_index = self.add_operand(i1_node.name, Operand.IOTYPE_INPUT)
-            np.array([input_operand_index], dtype=np.uint32).tofile(f)
-        elif i1_node.op == 'Const':
-            scalar = i1_node.attr['value'].tensor.float_val[0]
-            np.array([0], dtype=np.uint32).tofile(f)
-            input_operand_index = self.add_operand(i0_node.name, Operand.IOTYPE_INPUT)
-            np.array([input_operand_index], dtype=np.uint32).tofile(f)
-            np.array([1], dtype=np.uint32).tofile(f)
-            np.array([scalar], dtype=np.float32).tofile(f)
-        else:
-            np.array([0], dtype=np.uint32).tofile(f)
-            input_operand_index = self.add_operand(i0_node.name, Operand.IOTYPE_INPUT)
-            np.array([input_operand_index], dtype=np.uint32).tofile(f)
-            np.array([0], dtype=np.uint32).tofile(f)
-            input_operand_index = self.add_operand(i1_node.name, Operand.IOTYPE_INPUT)
-            np.array([input_operand_index], dtype=np.uint32).tofile(f)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([output_operand_index], dtype=np.uint32).tofile(f)
-
-
-    def dump_mathunary_to_file(self, node, f):
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-        i0_node = self.name_node_dict[node.input[0]]
-        np.array([self.op2code['MathUnary'], self.mathun2code[node.op]], dtype=np.uint32).tofile(f)
-        input_operand_index = self.add_operand(i0_node.name, Operand.IOTYPE_INPUT)
-        np.array([input_operand_index], dtype=np.uint32).tofile(f)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([output_operand_index],dtype=np.uint32).tofile(f)
-
-
-    def dump_avg_pool_to_file(self, node, f):
-        assert(node.op == 'AvgPool')
-        self.layer_number = self.layer_number + 1
-        self.converted_nodes.add(node.name)
-        node0 = self.name_node_dict[node.input[0]]
-        strides = node.attr['strides']
-
-        # Tensorflow do not support pooling strides in batch dimension and
-        # current native NN do not support pooling strides in channel dimension, added assert() here.
-        assert(strides.list.i[1]==strides.list.i[2])
-        assert(strides.list.i[0]==1)
-        assert(strides.list.i[3]==1)
-        strides = strides.list.i[1]
-        filter_node = node.attr['ksize']
-        input_name = node.input[0]
-
-        # Tensorflow do not support pooling ksize in batch dimension and channel dimension.
-        assert(filter_node.list.i[0]==1)
-        assert(filter_node.list.i[3]==1)
-        filter_height = filter_node.list.i[1]
-        filter_width = filter_node.list.i[2]
-
-        padding = node.attr['padding'].s.decode("utf-8")
-        np.array([self.op2code[node.op], strides, self.pool_paddings[padding], filter_height],
-                 dtype=np.uint32).tofile(f)
-
-        input_operand_index = self.add_operand(input_name, Operand.IOTYPE_INPUT)
-        output_operand_index = self.add_operand(node.name, Operand.IOTYPE_OUTPUT)
-        np.array([input_operand_index, output_operand_index],dtype=np.uint32).tofile(f)
-
-
-    def dump_layers_to_file(self, f):
-        for node in self.nodes:
-            if node.name in self.converted_nodes:
-                continue
-
-            # conv2d with dilation generates very complex nodes, so handle it in special
-            if self.in_conv2d_scope(node.name):
-                if node.op == 'Conv2D':
-                    self.dump_complex_conv2d_to_file(node, f)
-                continue
-            if self.in_dense_scope(node.name):
-                if node.op == 'MatMul':
-                    self.dump_dense_to_file(node, f)
-                continue
-
-
-            if node.op == 'Conv2D':
-                self.dump_simple_conv2d_to_file(node, f)
-                continue
-            if node.name in self.output_names:
-                input_name = self.id_different_scope_dict[node.name]
-                if TFConverter.get_scope_name(input_name)!=TFConverter.get_scope_name(node.name):
-                    continue
-            if node.op == 'AvgPool':
-                self.dump_avg_pool_to_file(node, f)
-            elif node.op == 'DepthToSpace':
-                self.dump_depth2space_to_file(node, f)
-            elif node.op == 'MirrorPad':
-                self.dump_mirrorpad_to_file(node, f)
-            elif node.op == 'Maximum':
-                self.dump_maximum_to_file(node, f)
-            elif node.op in self.mathbin2code:
-                self.dump_mathbinary_to_file(node, f)
-            elif node.op in self.mathun2code:
-                self.dump_mathunary_to_file(node, f)
-
-
-    def dump_operands_to_file(self, f):
-            operands = sorted(self.name_operand_dict.values())
-            for operand in operands:
-                #print('{}'.format(operand))
-                np.array([operand.index, len(operand.name)], dtype=np.uint32).tofile(f)
-                f.write(operand.name.encode('utf-8'))
-                np.array([operand.iotype, operand.dtype], dtype=np.uint32).tofile(f)
-                np.array(operand.dims, dtype=np.uint32).tofile(f)
-
-
-    def dump_to_file(self):
-        with open(self.outfile, 'wb') as f:
-            f.write(header.str.encode('utf-8'))
-            np.array([header.major, header.minor], dtype=np.uint32).tofile(f)
-            self.dump_layers_to_file(f)
-            self.dump_operands_to_file(f)
-            np.array([self.layer_number, len(self.name_operand_dict)], dtype=np.uint32).tofile(f)
-
-
-    def generate_name_node_dict(self):
-        for node in self.nodes:
-            self.name_node_dict[node.name] = node
-
-
-    def generate_output_names(self):
-        used_names = []
-        for node in self.nodes:
-            for input in node.input:
-                used_names.append(input)
-
-        for node in self.nodes:
-            if node.name not in used_names:
-                self.output_names.append(node.name)
-
-
-    def remove_identity(self):
-        self.id_different_scope_dict = {}
-        id_nodes = []
-        id_dict = {}
-        for node in self.nodes:
-            if node.op == 'Identity':
-                name = node.name
-                input = node.input[0]
-                id_nodes.append(node)
-                # do not change the output name
-                if name in self.output_names:
-                    self.name_node_dict[input].name = name
-                    self.name_node_dict[name] = self.name_node_dict[input]
-                    del self.name_node_dict[input]
-                    self.id_different_scope_dict[name] = input
-                else:
-                    id_dict[name] = input
-
-        for idnode in id_nodes:
-            self.nodes.remove(idnode)
-
-        for node in self.nodes:
-            for i in range(len(node.input)):
-                input = node.input[i]
-                if input in id_dict:
-                    node.input[i] = id_dict[input]
-
-
-    def generate_edges(self):
-        for node in self.nodes:
-            for input in node.input:
-                if input in self.edges:
-                    self.edges[input].append(node)
-                else:
-                    self.edges[input] = [node]
-
-
-    @staticmethod
-    def get_scope_name(name):
-        index = name.rfind('/')
-        if index == -1:
-            return ""
-        return name[0:index]
-
-
-    def in_conv2d_scope(self, name):
-        inner_scope = TFConverter.get_scope_name(name)
-        if inner_scope == "":
-            return False;
-        for scope in self.conv2d_scope_names:
-            index = inner_scope.find(scope)
-            if index == 0:
-                return True
-        return False
-
-
-    def in_dense_scope(self, name):
-        inner_scope = TFConverter.get_scope_name(name)
-        if inner_scope == "":
-            return False;
-        for scope in self.dense_scope_names:
-            index = inner_scope.find(scope)
-            if index == 0:
-                return True
-        return False
-
-    def generate_sub_block_op_scope_info(self):
-        # mostly, conv2d/dense is a sub block in graph, get the scope name
-        for node in self.nodes:
-            if node.op == 'Conv2D':
-                scope = TFConverter.get_scope_name(node.name)
-                # for the case tf.nn.conv2d is called directly
-                if scope == '':
-                    continue
-                # for the case tf.nn.conv2d is called within a scope
-                if scope + '/kernel' not in self.name_node_dict:
-                    continue
-                self.conv2d_scope_names.add(scope)
-            elif node.op == 'MatMul':
-                scope = TFConverter.get_scope_name(node.name)
-                # for the case tf.nn.dense is called directly
-                if scope == '':
-                    continue
-                # for the case tf.nn.dense is called within a scope
-                if scope + '/kernel' not in self.name_node_dict and scope.split('/Tensordot')[0] + '/kernel' not in self.name_node_dict:
-                    continue
-                self.dense_scope_names.add(scope.split('/Tensordot')[0])
-
-        # get the input name to the conv2d/dense sub block
-        for node in self.nodes:
-            scope = TFConverter.get_scope_name(node.name)
-            if scope in self.conv2d_scope_names:
-                if node.op == 'Conv2D' or node.op == 'Shape':
-                    for inp in node.input:
-                        if TFConverter.get_scope_name(inp) != scope:
-                            self.conv2d_scopename_inputname_dict[scope] = inp
-            elif scope in self.dense_scope_names:
-                if node.op == 'MatMul' or node.op == 'Shape':
-                    for inp in node.input:
-                        if TFConverter.get_scope_name(inp) != scope:
-                            self.dense_scopename_inputname_dict[scope] = inp
-            elif scope.split('/Tensordot')[0] in self.dense_scope_names:
-                if node.op == 'Transpose':
-                    for inp in node.input:
-                        if TFConverter.get_scope_name(inp).find(scope)<0 and TFConverter.get_scope_name(inp).find(scope.split('/')[0])<0:
-                            self.dense_scopename_inputname_dict[scope.split('/Tensordot')[0]] = inp
-
-
-    def run(self):
-        self.generate_name_node_dict()
-        self.generate_output_names()
-        self.remove_identity()
-        self.generate_edges()
-        self.generate_sub_block_op_scope_info()
-
-        if self.dump4tb:
-            self.dump_for_tensorboard()
-
-        self.dump_to_file()
-
-
-def convert_from_tensorflow(infile, outfile, dump4tb):
-    with open(infile, 'rb') as f:
-        # read the file in .proto format
-        graph_def = tf.GraphDef()
-        graph_def.ParseFromString(f.read())
-        nodes = graph_def.node
-
-    converter = TFConverter(graph_def, nodes, outfile, dump4tb)
-    converter.run()
diff --git a/tools/python/convert_header.py b/tools/python/convert_header.py
deleted file mode 100644
index 143f92c42e..0000000000
--- a/tools/python/convert_header.py
+++ /dev/null
@@ -1,26 +0,0 @@ 
-# Copyright (c) 2019
-#
-# 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
-# ==============================================================================
-
-str = 'FFMPEGDNNNATIVE'
-
-# increase major and reset minor when we have to re-convert the model file
-major = 1
-
-# increase minor when we don't have to re-convert the model file
-minor = 23