Commit Graph

1360 Commits

Author SHA1 Message Date
Alexander Alekhin
0a86ddc36c Merge pull request #19435 from l-bat:lb/onnx_normalize 2021-02-03 09:16:06 +00:00
Liubov Batanina
68eb54dc13 Added ONNX NormalizeL2 subgraph 2021-02-01 12:44:48 +03:00
Alexander Alekhin
b9dfffc237 Merge pull request #19057 from sl-sergei:changes_for_cuda_1d 2021-01-30 11:39:43 +00:00
Sergey Slashchinin
e2949c7d0a Align 3.4 branch with master 2021-01-29 23:48:08 +03:00
Alexander Alekhin
573812d0ff Merge pull request #19373 from l-bat:lb/tf_matmul_shared 2021-01-25 22:08:24 +00:00
Liubov Batanina
c12930cdde
Merge pull request #19372 from l-bat:lb/onnx_pads_calc
* Fixed bug with constant Div

* Supported constant mul and div for inputs with different shapes
2021-01-25 22:07:50 +00:00
Liubov Batanina
38a49f92ab Added shared weights for MatMul 2021-01-22 16:47:02 +03:00
Liubov Batanina
af9597f454 Added Mish ONNX subgraph 2021-01-15 14:01:48 +03:00
Alexander Alekhin
55f06b76f9 dnn: improve debugging of TensorFlow parsing errors 2020-12-26 15:31:30 +00:00
Alexander Alekhin
9b4adc9acb Merge pull request #19146 from alalek:dnn_openvino_2021.2.0 2020-12-17 19:50:36 +00:00
Vincent Rabaud
8391a23600 Optimize calls to std::string::find() and friends for a single char.
The character literal overload is more efficient. More info at:

http://clang.llvm.org/extra/clang-tidy/checks/performance-faster-string-find.html
2020-12-17 09:39:23 +01:00
Alexander Alekhin
28aab134db dnn(test): update tests for OpenVINO 2021.2 2020-12-17 07:53:35 +00:00
Alexander Alekhin
752cc26ad6 dnn: use OpenVINO 2021.2 defines
original commit: 4699d2ba0c
2020-12-16 22:22:17 +00:00
Alexander Alekhin
7631056b8a Merge pull request #19114 from alalek:issue_18937 2020-12-15 20:47:05 +00:00
Alexander Alekhin
c240355cc6 dnn(ocl): avoid mess FP16/FP32 in convolution layer 2020-12-15 08:51:24 +00:00
Alexander Alekhin
4b3d2c8834 dnn(ocl): fix gemm kernels with beta=0
- dst is not initialized, may include NaN values
- 0*NaN produces NaN
2020-12-15 00:58:43 +00:00
Alexander Alekhin
23c246882e Merge pull request #19071 from LupusSanctus:am/dnn_nearest_resize_3.4 2020-12-11 18:25:55 +00:00
Alexander Alekhin
d6a7f5e1e0 Merge pull request #19075 from alalek:dnn_fix_halide_build 2020-12-10 20:37:37 +00:00
Alexander Alekhin
d84a9484b7 dnn: fix build with Halide, skip tests with crashes 2020-12-10 18:23:24 +00:00
Anastasia Murzova
f2422ace7d Added TF nearest neighbour resize behaviour alignment
Relates to OpenCV issue #18721
2020-12-10 15:53:24 +03:00
Alexander Alekhin
8286d84fb1 Merge pull request #19061 from alalek:dnn_load_face_detector_with_external_protobuf 2020-12-09 18:03:52 +00:00
Alexander Alekhin
d7e936de5c dnn(caffe): add DetectionOutputParameter.clip to .proto file
- allow to load opencv_face_detector with external unpatched protobuf
2020-12-09 05:21:07 +00:00
Anastasia Murzova
a82c50eac2 Added TF resize_bilinear behaviour alignment
Fixes OpenCV issue #18721
2020-12-08 22:51:38 +03:00
Alexander Alekhin
00f36a3149 dnn: prefer to use v_fma() instead of v_c += v_a * v_b 2020-12-05 11:51:03 +00:00
Alexander Alekhin
2cf2456f4c dnn(test): skip unstable GatherMultiOutput OCL_FP16 test 2020-11-26 21:30:21 +00:00
Sergei Slashchinin
f4f462c50b
Merge pull request #18862 from sl-sergei:support_pool1d
Support for Pool1d layer for OpenCV and OpenCL targets

* Initial version of Pool1d support

* Fix variable naming

* Fix 1d pooling for OpenCL

* Change support logic, remove unnecessary variable, split the tests

* Remove other depricated variables

* Fix warning. Check tests

* Change support check logic

* Change support check logic, 2
2020-11-24 16:52:45 +00:00
Alexander Alekhin
0401d5920c Merge pull request #18845 from joegeisbauer:fix_reduce_mean_index_error 2020-11-23 17:03:47 +00:00
Joe
e05c2e0f1d Fix Reduce Mean error for MobileNets DNN
Fix for index error for Reduce Mean

Correct Reduce Mean indexing error
2020-11-20 11:17:02 -06:00
Julien
ac24a72e66
Merge pull request #18841 from JulienMaille:patch-2
Fixing dnn Resize layer for variable input size

* Fix onnx loading of resize/upsample layers for different opset

* group all DynamicResize tests

* cleaned up scales checks

* Simplify branching
2020-11-20 11:14:00 +00:00
Alexander Alekhin
9485113923 pre: OpenCV 3.4.13 (version++) 2020-11-17 21:50:30 +00:00
Alexander Alekhin
4f3130f562 Merge pull request #18145 from sl-sergei:fix_17914 2020-11-17 21:46:08 +00:00
Alexander Alekhin
f7e8dc770a Merge pull request #18834 from l-bat:update_reducemax 2020-11-17 21:14:10 +00:00
Alexander Alekhin
34909d97b4
Merge pull request #18840 from alalek:dnn_test_skip_myriad_gather_multi_output 2020-11-17 23:45:37 +03:00
Alexander Alekhin
474a67231c dnn(test): skip gather_multi_output test on Myriad 2020-11-17 19:52:07 +00:00
Liubov Batanina
72d06080c6 [ONNX] Added Reduce ops for batch and channel 2020-11-17 14:45:36 +03:00
Sergey Slashchinin
3cdf926454 disable Conv1d test on NGRAPH/MYRIAD 2020-11-17 14:33:39 +03:00
Alexander Alekhin
025a9647af Merge pull request #18830 from l-bat:issue_18785 2020-11-17 10:40:53 +00:00
Sergey Slashchinin
32e7ef8a3d Add fixes and tests for different layers 2020-11-17 13:39:32 +03:00
Sergei Slashchinin
2b82f8f12c
Merge pull request #18296 from sl-sergei:fix_16783
Fix loading issue for Faster RCNN model from #16783

* Add a reproducer with multi-output Gather

* Fix an issue with ONNX graph simplifier

* fix build

* Move checks to correct class

* Minor changes for better code appearence
2020-11-17 09:52:08 +00:00
Liubov Batanina
3a184ae677 [ONNX] Added handler for int32 tensors 2020-11-17 10:17:06 +03:00
Sergei Slashchinin
61144f935e
Merge pull request #18783 from sl-sergei:fix_conv1d
Add support for Conv1D on OpenCV backend

* Add support for Conv1D on OpenCV backend

* disable tests on other targets/backends

* Fix formatting

* Restore comment

* Remove unnecessary flag and fix test logic

* Fix perf test

* fix braces

* Fix indentation, assert check and remove unnecessary condition

* Remove unnecessary changes

* Add test cases for variable weights and bias

* dnn(conv): fallback on OpenCV+CPU instead of failures

* coding style
2020-11-13 22:22:10 +00:00
Alexander Alekhin
c3e7a23da5 Merge pull request #18752 from alalek:dnn_defines_openvino_2021.1.0 2020-11-08 18:09:52 +00:00
Alexander Alekhin
bed5debca6 dnn: use OpenVINO 2021.1 defines 2020-11-07 17:27:33 +00:00
Roman Kazantsev
bb5b628cce Use explicit opset of Unsqueeze from nGraph
backporting commit eb24575e2c
2020-11-06 22:29:35 +00:00
Alexander Alekhin
cdcf7e62f3 dnn(opencl): bypass unsupported fusion cases 2 2020-10-09 18:59:08 +00:00
Alexander Alekhin
718dd9f170 dnn(opencl): bypass unsupported fusion cases 2020-10-09 12:33:06 +00:00
Alexander Alekhin
e87a0baa4b dnn(test): enable tests from issue 17953 2020-10-08 20:27:03 +00:00
Alexander Alekhin
6da05f7086 dnn(test): update tests for OpenVINO 2021.1 2020-10-08 10:22:31 +00:00
Alexander Alekhin
a5f0fb6008 Merge pull request #18518 from alalek:backport_17993 2020-10-06 22:47:45 +00:00
Alexander Alekhin
037a72debd Merge pull request #18517 from alalek:backport_18031 2020-10-06 19:56:49 +00:00
Maksim Doronin
36f61f3879 [IE][VPU]: Refactor vpu configs
backported commit: 7fe87d9a5b
2020-10-05 20:27:52 +00:00
Ilya Churaev
aa11f7d8a3 Removed get_output_as_single_output_node method
backported commit: 5fd3d36fe8
2020-10-05 20:24:21 +00:00
Alexander Alekhin
a00fe15abd dnn: check for empty Net in .forward() 2020-10-05 06:23:47 +00:00
zhuqiang00099
a968eadbf1 fix darknet-relu bug in darknet_io.cpp 2020-10-02 06:16:38 +00:00
Alexander Alekhin
969b55036f Merge pull request #18438 from alalek:dnn_onnx_importer_error_reporting 2020-09-29 13:49:02 +00:00
Alexander Alekhin
c08f29c803 dnn(opencl): fix convolution kernel w/o bias with activation 2020-09-27 23:42:30 +00:00
Alexander Alekhin
e59793cc75 dnn: improve debugging of ONNX parsing errors 2020-09-27 23:04:48 +00:00
Alexander Alekhin
236ad4aeda Merge pull request #18441 from alalek:core_check_force_string_literals 2020-09-27 23:03:18 +00:00
Alexander Alekhin
233030e417 core: force check for string literals are used in the message 2020-09-27 06:37:44 +00:00
Tomoaki Teshima
48368dc9a1 loosen threshold for Mali 2020-09-27 00:37:52 +09:00
Alexander Alekhin
220b37144b Merge pull request #18395 from tomoaki0705:fixNativePow 2020-09-23 18:27:45 +00:00
Tomoaki Teshima
74c8ccb45b fix build error of kernel on Mali 2020-09-23 21:38:12 +09:00
Alexander Alekhin
4fa82809df ocl: avoid rescheduling of async kernels 2020-09-18 14:53:50 +00:00
Alexander Alekhin
3e3787ecb6 Merge pull request #18360 from tomoaki0705:fixClampFailure 2020-09-18 13:10:36 +00:00
Liubov Batanina
ebb528976f
Merge pull request #18353 from l-bat:issue_18350
* Fixed bug in ONNX Mul op

* Replaced node
2020-09-18 13:01:14 +00:00
Tomoaki Teshima
f77c2d700f add explicit cast for half 2020-09-18 21:04:24 +09:00
Sergei Slashchinin
fa953e4205
Merge pull request #18316 from sl-sergei:fix_18253
Fix loading of ONNX models with Resize operation with Opset 11 for newer versions of Pytorch

* Add reproducer for Resize operation from newer versions of Pytorch

* Fix loading of scales parameter for Resize layer

* Change check type for better diagnostic messages
2020-09-17 11:05:22 +00:00
Alexander Alekhin
4b24ddd70d Merge pull request #18317 from sl-sergei:restored_pr_17629 2020-09-13 12:51:41 +00:00
Shubham Singh
23e71d1aa2 fixes #17187 probably
Added Eltwise Layer Support
2020-09-11 18:53:42 +03:00
Liubov Batanina
b542a1804c Support global reduce ops 2020-09-09 11:56:20 +03:00
Alexander Alekhin
50ff40d684 pre: OpenCV 3.4.12 (version++) 2020-09-06 22:26:32 +00:00
Alexander Alekhin
fa11b98800 Merge pull request #18255 from alalek:backport_18243 2020-09-02 22:50:21 +00:00
Alexander Alekhin
64c67a93d3 Merge pull request #18246 from YashasSamaga:dnn-permute-fix-unwanted-ocl-init 2020-09-02 21:15:19 +00:00
Alexander Alekhin
1f2c83845d backport: checks and fixes from static code analyzers results
original commit: 71f665bd8c
2020-09-02 19:05:47 +00:00
Liubov Batanina
2349a09736 Support Reshape with zero dim 2020-09-02 11:58:43 +03:00
YashasSamaga
1df533c914 fix typo in fusion tests 2020-09-02 14:25:36 +05:30
YashasSamaga
44bf748479 do not allocate UMat in non-OpenCL targets 2020-09-02 12:18:41 +05:30
Alexander Alekhin
792722865f Merge pull request #17919 from zhaoyue-zephyrus:flownet2_with_anysize 2020-08-26 11:43:17 +00:00
Sergei Slashchinin
9aa401a7d0
Merge pull request #17978 from sl-sergei:fix_17516_17531
* Fix ONNX loading in issues opencv#17516, opencv#17531

* Add tests for Linear and Matmul layers

* Disable tests for IE versions lower than 20.4

* Skip unstable tests with OpenCL FP16 on Intel GPU

* Add correct test filtering for OpenCL FP16 tests
2020-08-26 10:15:59 +00:00
Liubov Batanina
d392b11dfb Supported ONNX Pow op 2020-08-24 11:20:18 +03:00
Liubov Batanina
ad63d24dba
Merge pull request #18096 from l-bat:update_onnx_importer
* Added ReduceSum to ONNX importer

* Fix comments

* Fix Mul
2020-08-14 16:49:42 +00:00
Alexander Alekhin
3b5813c035 Merge pull request #18078 from l-bat:fix_matmul 2020-08-14 13:46:46 +00:00
Liubov Batanina
339b963e6b Fix MatMul and Add axes 2020-08-14 11:18:58 +03:00
Yashas Samaga B L
2171cae8ff
Merge pull request #17976 from YashasSamaga:dnn-fusion-tests-fix-ocl
dnn: add exhaustive fusion tests, enable more eltwise fusions

* add eltwise fusion tests, enable more eltwise fusions

* merge weighted eltwise tests with eltwise tests
2020-08-13 10:55:41 +00:00
Liubov Batanina
f3cebb3e1b
Merge pull request #18077 from l-bat:reduce_sum
* Supported ReduceSum op

* Skip test
2020-08-12 14:32:16 +00:00
zhaoyue-zephyrus
e231be86b7 support flownet2 with arbitary input size
revise default proto to match the filename in documentations

fix a bug

beautify python codes

fix bug

beautify codes

add test samples with larger/smaller size

remove unless code

using bytearray without creating tmp file

remove useless codes
2020-08-12 00:50:58 +08:00
Liubov Batanina
6226ea0085 Fix bug in ONNX Gather op 2020-08-06 15:47:34 +03:00
Vadim Pisarevsky
1537ecd931 * added depth-wise convolution; gives ~20-30% performance improvement in MobileSSD networks
* hopefully, eliminated compile warnings, errors, as well as failure in one test

* * fixed a few typos
* decreased buffer size in some cases
* added more optimal im2row branch in the case of 1x1 convolutions
* tuned fastConv to reduce the number of passes over arrays

backport of commit 77b01deb80
2020-08-04 17:34:48 +00:00
Alexander Alekhin
5b5c42d2c7 Merge pull request #18027 from dkurt:dnn_backport_ngraph 2020-08-04 16:24:11 +00:00
Alexander Alekhin
161890dad4 Merge pull request #18017 from danielenricocahall:add-relu-to-darknet 2020-08-04 16:17:07 +00:00
danielenricocahall
8457e471fd add relu as activation option in darknet
add relu option

add relu as activation option in darknet

simplify the setParams if-else ladder

add relu as activation option in darknet

correct activation_param type

format

format

add relu as activation option in darknet

spacing

spacing

add relu as activation option in darknet
2020-08-03 19:19:35 -04:00
Ilya Churaev
246de2b7f5 Replaced copy_with_new_args to clone_with_new_inputs 2020-08-03 23:08:29 +03:00
Ilya Churaev
e8c26963e9 Fixed removing is_parameter, is_constant, is_output 2020-08-03 23:08:22 +03:00
Alexander Alekhin
a28533933f Merge pull request #17998 from dkurt:dnn_fix_ngraph 2020-08-03 18:23:11 +00:00
Liubov Batanina
d695208727
Merge pull request #17967 from l-bat:non_const_weights_for_conv
* Supported convolution with non-const weights

* Fix opencl blobs

* Update tests
2020-08-03 18:02:49 +00:00
Alexander Alekhin
65b02cc8f2
Merge pull request #17742 from SoheibKadi/DetectionOutput_layer_doc
Adding comment from source code to DetectionOutputLayer class documentation
2020-08-03 17:17:04 +03:00
Alexander Alekhin
1c8ee3f957
Merge pull request #17885 from alalek:dnn_ocl_slice_update
DNN: OpenCL/slice update

* dnn(ocl/slice): make slice kernel VTune friendly

- more unique names
- inline code of copy functions

* dnn(ocl/slice): prefer to spawn more work groups

- even in case with 1D copy
- perf improvement up to 2x of kernel time (due to changed configuration 128x1x1 => 128x32x1)

* dnn(ocl/slice): cache kernel exec info
2020-08-03 14:13:34 +00:00
Dmitry Kurtaev
cf8f65d806 Do not use size_t for nGraph layers 2020-08-02 20:50:44 +03:00
kadi soheib
6bed5c181b Corrected Comment as requested by reviewer. 2020-07-31 23:43:38 +03:00
Sinitsina
0ac2f0e04c mish_functor_update 2020-07-23 09:02:00 +03:00
Liubov Batanina
a35d4f9029 Support Gather for variable inputs 2020-07-20 14:02:45 +03:00
Alexander Alekhin
b2ebd37ee2 Merge pull request #17856 from alalek:dnn_openvino_2020.4.0 2020-07-16 20:08:00 +00:00
Alexander Alekhin
09f24a851e Merge pull request #17764 from alalek:issue_17762 2020-07-16 15:27:21 +00:00
Alexander Alekhin
81e027eef7 dnn: fix OpenCL implementation of Slice layer 2020-07-16 04:33:52 +00:00
Alexander Alekhin
1c371d07b5 dnn(test): adjust tests for OpenVINO 2020.4 2020-07-15 23:47:40 +00:00
Alexander Alekhin
55e8549839 dnn: eliminate IE deprecation warning 2020-07-15 23:39:06 +00:00
Alexander Alekhin
435b6df989 dnn: use OpenVINO 2020.4 defines
original commit: 2813aa7eb9
2020-07-15 20:13:40 +00:00
Dmitry Kurtaev
cc584760d3 Fix TensorFlow->ONNX imports 2020-07-15 14:36:13 +03:00
Alexander Alekhin
5cb8619eca dnn(ie): enable KEY_CPU_THREADS_NUM for Windows 2020-07-10 14:29:21 +00:00
Alexander Alekhin
950a916952 Merge pull request #17752 from YashasSamaga:generalize-concat-fusion-3.4 2020-07-07 10:36:02 +00:00
Alexander Alekhin
d62e0a3695 Merge pull request #17743 from alalek:test_17666 2020-07-06 22:36:35 +00:00
Alexander Alekhin
d5713c657b dnn(slice): disable buggy OCV/OCL implementation 2020-07-06 21:36:19 +00:00
Alexander Alekhin
99c4b76a6d dnn(test): add YOLOv4-tiny tests 2020-07-06 21:36:19 +00:00
Ilya Churaev
d69a7a3bbf Fixed header paths for some nGraph ops
* Added dependency on IE version

backport of commit: 992c908b56
2020-07-06 19:51:57 +00:00
YashasSamaga
b7eec216e9 generalize axis for concat fusion 2020-07-04 18:57:28 +05:30
kadi soheib
17c430da88 Updated comment. 2020-07-04 06:37:59 +03:00
kadi soheib
96a501c08b Adding comment from source code to documentation. 2020-07-04 06:37:58 +03:00
Liubov Batanina
65dbbf712d
Merge pull request #17733 from l-bat:tiny_yolov4
* Supported yolov4-tiny

* Added comments
2020-07-03 18:07:08 +00:00
Ilya Lavrenov
e58ce89b10 Conditional compilation for IR v7 support
backported commit 86905754e4
2020-06-25 10:05:14 +00:00
Alexander Alekhin
2e165053ef Merge pull request #17624 from dkurt:dnn_optimize_mish 2020-06-23 18:43:51 +00:00
Ilya Lavrenov
ad5e70f94e Conditional compilation for network reader
origibal commit: 63e92cccf2
2020-06-23 14:15:52 +03:00
Dmitry Kurtaev
8e3f5fb209 Remove deprecated Inference Engine CPU extensions 2020-06-23 10:10:29 +03:00
Dmitry Kurtaev
1491934d17 Optimize Mish for CPU backend 2020-06-22 23:27:47 +03:00
Liubov Batanina
85c0c8c7ed Disabling dafault NMS in yolo layer 2020-06-19 14:34:13 +03:00
Liubov Batanina
d93b6be3cc Changed StridedSlice to VariadicSplit in Region layer 2020-06-17 10:02:53 +03:00
Ilya Lavrenov
676b818d6a Removed plugin dispatcher
backport of commit 74113737f3
2020-06-15 18:03:14 +00:00
Alexander Alekhin
1644af841b Merge pull request #17536 from ilya-lavrenov:remove-error-listener 2020-06-12 22:18:04 +00:00
Ilya Lavrenov
9697e3ac24 Removed error lisneter usage 2020-06-12 20:29:11 +00:00
Alexander Alekhin
d2a9efd039 dnn: use OpenVINO 2020.3 defines 2020-06-12 20:24:08 +00:00
Maksim Shabunin
9096b1c768 dnn/NGraph: added nullptr checks 2020-06-10 13:48:24 +03:00
Alexander Alekhin
a43e3bebe6 pre: OpenCV 3.4.11 (version++) 2020-06-08 18:46:27 +00:00
Alexander Alekhin
08d1c54364 Merge pull request #17469 from l-bat:fix_virtual_try_on 2020-06-04 10:32:30 +00:00
Liubov Batanina
5ffc5bca7c Fixed virtual try on sample 2020-06-04 09:41:24 +03:00
Alexander Alekhin
d60524ecfa Merge pull request #17447 from alalek:dnn_ie_extract_layers 2020-06-02 12:40:06 +00:00
Alexander Alekhin
be1a121d0d dnn(ie): fix layers extraction 2020-06-01 21:57:39 +00:00
Alexander Alekhin
8de176988d Merge pull request #17439 from alalek:fix_dnn_test_required_file 2020-05-31 19:51:56 +00:00
Alexander Alekhin
79c5d07abe dnn(test): file 'dnn/efficientdet-d0.pb' is optional 2020-05-31 16:18:37 +00:00
Alexander Alekhin
740f6628ec Merge pull request #17419 from YourButterfly:fix_divide_zero_in_darknet_3_4 2020-05-29 17:32:48 +00:00
YourButterfly
f7daa9e4f5 chk divide 0 2020-05-29 09:54:46 +08:00
Alexander Alekhin
e58e545584 Merge pull request #17392 from alalek:dnn_test_yolov4 2020-05-28 22:52:21 +00:00
Alexander Alekhin
319db07b6b Merge pull request #17384 from dkurt:efficientdet 2020-05-28 22:48:52 +00:00
Dmitry Kurtaev
d9bada9867 dnn: EfficientDet 2020-05-28 17:23:42 +03:00
Alexander Alekhin
6b89154afd dnn(test): add YOLOv4 tests 2020-05-28 13:27:40 +00:00
Alexander Alekhin
7d9e1be588 Merge pull request #17388 from alalek:dnn_update_network_dump 2020-05-27 07:55:32 +00:00
Liubov Batanina
ba3cf47600
Merge pull request #17386 from l-bat:tf_clamp_subgraph
* Added ClipByValue subgraph

* Return const nodes
2020-05-26 19:01:47 +00:00
Liubov Batanina
b236f10792 Switch ngraph::op::v1::Multiply to v0 2020-05-26 16:59:50 +03:00
Alexander Alekhin
f0bef94a03 dnn: update network dump code, include ngraph serialization 2020-05-26 12:49:22 +00:00
Alexander Alekhin
73aa5f567b dnn: *_DENORMALS_ZERO_MODE is defined for SSE3 2020-05-25 17:55:36 +00:00
Alexander Alekhin
a9b030591b Merge pull request #17295 from dkurt:dnn_fusion_ftz 2020-05-22 18:49:43 +00:00
Liubov Batanina
d991c22090
Merge pull request #16575 from l-bat:flownet2
Support FlowNet2 model

* Support DataAugmentation layer

* Fix warnings

* Fix comments

* Support Correlation layer

* TEST

* Support Correlation layer

* Supported Accum and FlowWarp layers

* Supported ChannelNorm layer

* Supported Resample with inputs.size() > 1

* Fixed comments

* Refactoring

* Added tests

* Add resample test

* Added asserts in resize layer

* Updated DataAugmentation layer

* Update convolution layer

* Refactoring

* Fix data augmentation layer

* Fix caffe importer

* Fix resize

* Switch to Mat ptr

* Remove useless resize type

* Used ResizeLayer in Accum

* Split ChannelNormLayer

* Delete duplicate assert

* Add sample

* Fix sample

* Added colormap
2020-05-19 12:29:50 +00:00
Dmitry Kurtaev
b5035ce991 Increase test threshold for YOLOv3 on OCL FP16 2020-05-18 19:06:15 +03:00
Dmitry Kurtaev
68d59a2913 Flush to zero Convolution denormal weights 2020-05-15 23:44:34 +03:00
Alexander Alekhin
c8689d9d0a Merge pull request #17288 from dkurt:dnn_tf_resize_down 2020-05-14 07:08:41 +00:00
Alexander Alekhin
bcf96d637e Merge pull request #17284 from dkurt:dnn_bn_fusion 2020-05-14 06:57:36 +00:00
Dmitry Kurtaev
b4a6aa335d TensorFlow bilinear resize downscale 2020-05-13 23:59:20 +03:00
Dmitry Kurtaev
df305e83fa Fix BatchNorm reinitialization after fusion 2020-05-13 22:15:36 +03:00
Liubov Batanina
b27ae9c63b Switch v1::Multiply to v0::Multiply 2020-05-12 16:05:03 +03:00
Liubov Batanina
79f8b7fd73
Merge pull request #17233 from l-bat:onnx_bn
* Added ONNX BatchNorm subgraph

* Move removing constant inputs to addConstantNodesForInitializers

* Added initializers to ONNXGraphWrapper
2020-05-12 15:33:57 +03:00
Alexander Alekhin
066259b656
Merge pull request #17118 from l-bat/concat_3d
Added NDHWC Concat support for TensorFlow

* Supported TF concat 3d

* Skip myriad

* Fix test
2020-05-09 22:24:06 +03:00
Dmitry Kurtaev
8b13b85c5e dnn: Slice with variable input shapes 2020-05-05 13:35:17 +03:00
Liubov Batanina
a5696da9ec
Merge pull request #17185 from l-bat:yolo_v4
* Support Yolov4

* Skip Mish on OpenVINO 2020.2

* Revert Mish

* Refactoring
2020-04-30 16:53:44 +03:00
Alexander Alekhin
b805115c1a dnn(test): update skip tests on Win32 configuration 2020-04-29 20:02:13 +00:00
Dmitry Kurtaev
25ec4ce6f1 PReLU from Tensorflow 2020-04-28 00:01:21 +03:00
Alexander Alekhin
3f38edfd61 Merge pull request #17161 from alalek:cmake_protobuf_fix 2020-04-27 12:24:37 +00:00
Alexander Alekhin
9181ecfc7b cmake: fix protobuf handling 2020-04-27 02:11:19 +00:00
Liubov Batanina
4bf94cb5d1 Fix test 2020-04-26 20:42:11 +03:00
Alexander Alekhin
288fa70ed9 dnn(protobuf): backport AllowUnknownField(), SetRecursionLimit()
- limit recursion in SkipField*() calls
2020-04-25 20:45:43 +00:00
Alexander Alekhin
2df978b8e7 Merge pull request #17134 from alalek:dnn_ie_avoid_conversion_to_legacy 2020-04-23 22:39:45 +00:00
Alexander Alekhin
10808ccbb4 Merge pull request #17129 from alalek:dnn_myriad_tests 2020-04-23 19:41:15 +00:00
Alexander Alekhin
f756923271 dnn(ie): avoid conversion to legacy CNNNetworkImpl 2020-04-23 19:11:33 +00:00
Alexander Alekhin
83c4378d5d dnn(test): skip failed NGRAPH/MYRIAD tests 2020-04-22 15:00:23 +00:00
Liubov Batanina
1c1762d3f6 Skip myriad 2020-04-22 09:52:20 +03:00
ashishiva3@gmail.com
e0ac0cfbe2 add fused batchNorm Upsample 2020-04-22 08:24:17 +05:30
Alexander Alekhin
1a17f402a4 Merge pull request #17030 from ashishkrshrivastava:onnximporter 2020-04-21 21:02:41 +00:00
Liubov Batanina
aa08900ac8 Supported TF concat 3d 2020-04-21 15:15:22 +03:00
AshihsKrShrivastava
d37180a2c4 modification for upsample node fused from unfused Resize subgraph 2020-04-21 15:03:00 +05:30
Liubov Batanina
8badf7f354
Merge pull request #17112 from l-bat:ie_region
* Support nGraph Region

* Support region since OpenVINO 2020.2

* Skip myriad
2020-04-21 09:26:58 +00:00
Dmitry Kurtaev
908bf935f7 Flexible inputs for OpenVINO IR models 2020-04-18 20:00:22 +03:00
Alexander Alekhin
c007228d0f Merge pull request #17075 from alalek:dnn_updates_from_openvino_2020.2 2020-04-16 12:19:58 +00:00
Dmitry Kurtaev
5e5385a814 Remove NodeVector for nGraph 2020-04-16 11:13:49 +03:00
Alexander Alekhin
ef68cc3d5f dnn: use OpenVINO 2020.2 defines
original commit: 45263d7642
2020-04-15 11:14:43 +00:00
Alexander Alekhin
f6de25b6cb Merge pull request #17060 from dkurt:dnn_align_ie_pool 2020-04-14 15:15:44 +00:00
Ilya Lavrenov
91b0100287 Fixed compilation when NN builder is not built 2020-04-14 15:05:01 +03:00
Dmitry Kurtaev
870a775d7d Align DLIE and nGraph average pooling layers 2020-04-14 13:54:54 +03:00
Dmitry Kurtaev
d3f9ad1145 Enable ONNX SSD from https://github.com/amdegroot/ssd.pytorch 2020-04-13 15:12:27 +03:00
AshihsKrShrivastava
bef6b6282c ReflecitonPad2d and ZeroPad2d Subgraph fusion added 2020-04-11 07:14:05 +05:30
Alexander Alekhin
06bf845783 Merge pull request #16979 from dkurt:dnn_fused_resize_conv 2020-04-08 13:29:51 +00:00
Dmitry Kurtaev
21ed892489 Fix Elu import from ONNX 2020-04-08 12:09:45 +03:00
Alexander Alekhin
adf54d41d5 Merge pull request #16905 from dkurt:dnn_tf2_keras 2020-04-07 14:59:42 +00:00
Liubov Batanina
734771418e
Merge pull request #16840 from l-bat:matmul_inputs
* Supported FullyConnected layer with two inputs

* Skipped test

* Fix conditions

* Added OpenCL support

* Supported ReduceMean3D

* Supported Expand layer

* Fix warning

* Added Normalize subgraph

* refactoring

* Used addLayer

* Fix check

* Used addLayer

* Skip failed test

* Added normalize1 subgraph

* Fix comments
2020-04-07 14:12:18 +00:00
Alexander Alekhin
ce5626db45 Merge pull request #16878 from dkurt:fix_16877 2020-04-06 18:14:35 +00:00
Liubov Batanina
a448d3a6aa Fix ENet test with OpenVINO 2020.2 2020-04-06 17:29:43 +03:00
Dmitry Kurtaev
b36eba7fab Support FusedResizeAndPadConv2D from TensorFlow 2020-04-04 16:02:17 +03:00
Dmitry Kurtaev
8574a757f9 Case sensitive dnn layers types 2020-04-04 15:03:56 +03:00
Alexander Alekhin
e661ad2a67 eliminate build warnings 2020-03-27 11:39:07 +00:00
Dmitry Kurtaev
7e4b2057f2 Import TF2.0 network from Keras 2020-03-25 15:34:28 +03:00
Alexander Alekhin
e8c7d61752 Merge pull request #16817 from dkurt:dnn_onnx_lstm 2020-03-23 16:22:17 +00:00
Dmitry Kurtaev
467c3ef0ac Add checks for LSTM initial h and c 2020-03-23 16:28:55 +03:00
Dmitry Kurtaev
8433620295 Bidirectional LSTM 2020-03-22 00:56:48 +03:00
Pavel Rojtberg
66cf55ea1f dnn: expose only float variant of NMSBoxes for bindings
the float variant was always shadowed by the int version as
Rect2d is implicitly convertible to Rect.
This swaps things which is fine, as the vector of boxes was always
copied and the computation was done in double.
2020-03-19 12:36:35 +01:00