62fc7e3035
dnn_backend_native_layer_mathunary: add cosh support
...
Signed-off-by: Ting Fu <ting.fu@intel.com>
4 years ago
91b4037101
dnn_backend_native_layer_mathunary: add sinh support
...
Signed-off-by: Ting Fu <ting.fu@intel.com>
4 years ago
13f5613e68
dnn_backend_native_layer_mathunary: add atan support
...
It can be tested with the model generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.atan(x)
x2 = tf.divide(x1, 3.1416/4) # pi/4
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
461485feac
dnn_backend_native_layer_mathunary: add acos support
...
It can be tested with the model generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.acos(x)
x2 = tf.divide(x1, 3.1416/2) # pi/2
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
486c0c419d
dnn_backend_native_layer_mathunary: add asin support
...
It can be tested with the model generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.asin(x)
x2 = tf.divide(x1, 3.1416/2) # pi/2
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
0b182ff66d
tools/target_dec_fuzzer: Adjust threshold for lagarith
...
Fixes: Timeout (3minute 49 sec -> 3sec)
Fixes: 22020/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_LAGARITH_fuzzer-5708544679870464
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years ago
d3747f4431
tools/target_dem_fuzzer: Use file extensions listed in input formats
...
This should make it easier for the fuzzer to fuzz formats being detected only by
file extension and thus increase coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years ago
22d0860c13
dnn_backend_native_layer_mathunary: add tan support
...
It can be tested with the model generated with below python scripy
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.multiply(x, 0.78)
x2 = tf.tan(x1)
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
88fb494f42
dnn_backend_native_layer_mathunary: add cos support
...
It can be tested with the model generated with below python scripy
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.multiply(x, 1.5)
x2 = tf.cos(x1)
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
0b6d3f0d83
dnn_backend_native_layer_mathunary: add sin support
...
It can be tested with the model file generated with below python scripy:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.multiply(x, 3.14)
x2 = tf.sin(x1)
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo Yejun <yejun.guo@intel.com>
4 years ago
3371d0611f
tools/target_dec_fuzzer: enable mjpeg for tiff or tdsc
...
This is needed for fuzzing tiff/tdsc and should increase coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
3e651eeac4
tools/target_dem_fuzzer: Implement AVSEEK_SIZE
...
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
f73cc61bf5
dnn_backend_native_layer_mathunary: add abs support
...
more math unary operations will be added here
It can be tested with the model file generated with below python scripy:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpeg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.subtract(x, 0.5)
x2 = tf.abs(x1)
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
6d4fdb4f5a
tools/target_dec_fuzzer: Adjust max_pixels for AV_CODEC_ID_HAP
...
Fixes: Timeout (170sec -> 6sec)
Fixes: 20956/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HAP_fuzzer-5713643025203200
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
d6824ef905
tools/target_dec_fuzzer: Reduce maxpixels for HEVC
...
high resolutions with only small blocks appear to be rather
slow with the fuzzer + sanitizers.
A solution which makes this run faster is welcome.
Fixes: Timeout (did not wait -> 17sec)
Fixes: 21006/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HEVC_fuzzer-6002552539971584
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
05d364dccc
tools/target_dec_fuzzer: Do not test AV_CODEC_FLAG2_FAST with AV_CODEC_ID_H264
...
This combination skips allocating large padding which can read out of array
Fixes: 20978/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_H264_fuzzer-5746381832847360
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
ea980d4162
fate: add tests for h264 and vp9 video enc parameters export
5 years ago
bf80725352
lavc: rename bsf.h to bsf_internal.h
...
This will allow adding a public header named bsf.h
5 years ago
4f54982773
tools/target_dec_fuzzer: Adjust threshold for PNG and APNG
...
Fixes: Timeout (84sec -> 2sec)
Fixes: 21127/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_APNG_fuzzer-5098412367413248
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
71e28c5422
dnn/native: add native support for minimum
...
it can be tested with model file generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
x1 = tf.minimum(0.7, x)
x2 = tf.maximum(x1, 0.4)
y = tf.identity(x2, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
d817b57d36
tools: fix const specifier for AVInputFormat
...
Signed-off-by: Josh de Kock <josh@itanimul.li>
5 years ago
8ce9d88f93
dnn/native: add native support for divide
...
it can be tested with model file generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
z1 = 2 / x
z2 = 1 / z1
z3 = z2 / 0.25 + 0.3
z4 = z3 - x * 1.5 - 0.3
y = tf.identity(z4, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
ef79408e97
dnn/native: add native support for 'mul'
...
it can be tested with model file generated from above python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
z1 = 0.5 + 0.3 * x
z2 = z1 * 4
z3 = z2 - x - 2.0
y = tf.identity(z3, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
6aa7e07e7c
dnn/native: add native support for 'add'
...
It can be tested with the model file generated with below python script:
import tensorflow as tf
import numpy as np
import imageio
in_img = imageio.imread('input.jpg')
in_img = in_img.astype(np.float32)/255.0
in_data = in_img[np.newaxis, :]
x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in')
z1 = 0.039 + x
z2 = x + 0.042
z3 = z1 + z2
z4 = z3 - 0.381
z5 = z4 - x
y = tf.math.maximum(z5, 0.0, name='dnn_out')
sess=tf.Session()
sess.run(tf.global_variables_initializer())
graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out'])
tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False)
print("image_process.pb generated, please use \
path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n")
output = sess.run(y, feed_dict={x: in_data})
imageio.imsave("out.jpg", np.squeeze(output))
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
39962072a8
tools: stop using deprecated av_codec_next()
...
Signed-off-by: Josh de Kock <josh@itanimul.li>
5 years ago
2db37bf4cd
tools/target_dec_fuzzer: Adjust threshold for zerocodec
...
Fixes: Timeout (147sec -> 1sec)
Fixes: 20764/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_ZEROCODEC_fuzzer-5068274603917312
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
8dee1d7a30
tools/target_dec_fuzzer: Adjust threshold for screenpresso
...
Fixes: Timeout (332 -> 21 sec)
Fixes: 20280/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_SCREENPRESSO_fuzzer-6238663432470528
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
ffa1561608
dnn_backend_native_layer_mathbinary: add sub support
...
more math binary operations will be added here
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
5 years ago
32522b5307
tools/target_dec_fuzzer: limit per frame samples for APE
...
APE in its highest compression mode is really slow so even one frame
of millions of samples takes a long time
Fixes: Timeout (too long -> 3sec)
Fixes: 19937/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_APE_fuzzer-5751668818051072
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
48b6947821
tools/target_dec_fuzzer: Add threshold for ALS
...
Fixes: Timeout (253sec -> 16sec)
Fixes: 18668/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_ALS_fuzzer-6227155369590784
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
04e524c34b
tools/target_dec_fuzzer: Add threshold for IFF_ILBM
...
Fixes: Timeout (32 -> 1sec)
Fixes: 20138/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IFF_ILBM_fuzzer-5634665251864576
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Reviewed-by: Peter Ross <pross@xvid.org>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
cc7bf7e05c
tools/target_dec_fuzzer: Sort threshold list alphabetically
...
This also removes the comments as they are hard to maintain
together with sorted lists
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
5f7727e1c9
tools/target_dec_fuzzer: Use codec_tags list
...
This should make it much quicker for the fuzzer to test
real relevant codec_tags
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
00447b6f52
tools/target_dec_fuzzer: Also Fuzz with CPU optimizations disabled
...
This should improve coverage of *_c()
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
4b733a7f5f
tools/target_dec_fuzzer: Fuzz private options of AC3/E-AC3
...
This should improve AC-3 coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
48083f1890
tools/target_dec_fuzzer: Fuzz idct_algo value
...
This should improve coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
029cc9883f
tools/target_dec_fuzzer: Adjust threshold for RASC
...
Fixes: Timeout(35sec -> 4sec)
Fixes: 19289/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RASC_fuzzer-5676526398078976
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
c1411a1132
tools/target_dec_fuzzer: Stop negative block_align and sampling rate
...
These are checked for early in avcodec_open2() and do not really test the decoder
but instead waste resources which could be better spend fuzzing the actual decoder
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
5a8877da35
tools/target_dec_fuzzer: Adjust threshold for dst
...
Fixes: Timeout (400sec -> 14sec)
Fixes: 18989/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_DST_fuzzer-5175008116867072
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
e52070e89c
convert_from_tensorflow.py: add support when kernel size is 1*1 with one input/output channel (gray image)
...
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
5 years ago
e3dddf2142
tools/target_dec_fuzzer: Also fuzz request_channel_layout
...
This should improve coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
964eb754b4
tools: add a fuzzer tool for bitstream filters
...
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>
5 years ago
5ac8675cb1
tools/target_dec_fuzzer: Support setting AV_CODEC_FLAG2_FAST
...
This should improve coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
3ae87bb3c1
tools/target_dec_fuzzer: Support fuzzing error detection
...
This should increase coverage
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
e21ec54fdf
tools/target_dec_fuzzer: Adjust threshold for VP9
...
The threshold is chosen so that the worse frames would together not take
excessive time.
A better solution is welcome!
Fixes: Timeout (308sec ->102ms)
Fixes: 18314/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VP9_fuzzer-5701689176227840
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
de1d3b0925
tools/target_dec_fuzzer: Adjust threshold for smacker
...
Fixes: Timeout (65sec -> 0.5sec)
Fixes: 18072/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_SMACKER_fuzzer-5722709366931456
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
8ff432eb44
tools/enum_options: replace the deprecated API
...
Signed-off-by: Zhong Li <zhongli_dev@126.com>
5 years ago
f52e15df0f
tools/probetest: replace the deprecated API
...
Signed-off-by: Zhong Li <zhongli_dev@126.com>
5 years ago
fd3ee7a92e
tools/target_dec_fuzzer: Adjust threshold for G2M
...
G2M allows large images from small input and also reallocates
multiple buffers on any resolution change.
Fixes: Timeout (22sec -> 5sec)
Fixes: 18022/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_G2M_fuzzer-5089192530411520
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years ago
dff39ea9f0
dnn: add tf.nn.conv2d support for native model
...
Unlike other tf.*.conv2d layers, tf.nn.conv2d does not create many
nodes (within a scope) in the graph, it just acts like other layers.
tf.nn.conv2d only creates one node in the graph, and no internal
nodes such as 'kernel' are created.
The format of native model file is also changed, a flag named
has_bias is added, so change the version number.
Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
5 years ago
Ting Fu
Michael Niedermayer
Anton Khirnov
Guo, Yejun
Josh de Kock
James Almer
Zhong Li