mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-29 13:47:32 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity
29,912 Commits 6 Branches 126 Tags 3 GiB
28c0cd8460
Commit Graph

750 Commits

Author SHA1 Message Date
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
Alexander Alekhin
06bff34a6b Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-05-12 19:24:28 +00:00
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
Alexander Alekhin
09799402f9 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-05-06 19:53:51 +00:00
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
Alexander Alekhin
c722625f28 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-28 16:53:19 +00:00
Dmitry Kurtaev
25ec4ce6f1 PReLU from Tensorflow 2020-04-28 00:01:21 +03:00
Liubov Batanina
4bf94cb5d1 Fix test 2020-04-26 20:42:11 +03:00
Alexander Alekhin
152e6476d9 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-24 19:02:39 +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
bfcc136dc7 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-21 21:32:51 +00:00
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
ca9756f6a1 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-13 20:00:12 +00:00
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
1537a5f253 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-09 19:19:20 +00:00
Alexander Alekhin
06bf845783 Merge pull request #16979 from dkurt:dnn_fused_resize_conv 2020-04-08 13:29:51 +00:00
Alexander Alekhin
b8579f12be Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-04-08 10:19:09 +00:00
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
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
Dmitry Kurtaev
7e4b2057f2 Import TF2.0 network from Keras 2020-03-25 15:34:28 +03:00
Alexander Alekhin
0b4c101e8a Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-23 16:54:35 +00:00
Dmitry Kurtaev
467c3ef0ac Add checks for LSTM initial h and c 2020-03-23 16:28:55 +03:00
YashasSamaga
2aeb32d2d1 fix segfaults, support bias in untrainable mode, support batches in untrainable mode 2020-03-22 22:18:52 +05:30
Dmitry Kurtaev
8433620295 Bidirectional LSTM 2020-03-22 00:56:48 +03:00
Alexander Alekhin
4cdb4652cf Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-20 21:41:15 +00:00
Dmitry Kurtaev
8d69dbdf49 LSTM from ONNX works 2020-03-17 22:05:57 +03:00
Dmitry Kurtaev
14da5ec311 LSTM scalar 2020-03-17 22:01:49 +03:00
Liubov Batanina
718d7e4b04
Merge pull request #16715 from l-bat:slice_onnx 
* Support Slice layer with multiple inputs

* Add test

* Supported Resize from PyTorch

* Rewrite test

* Remove Cast layer (supported in #16735)

* Support ConstantOfShape

* Fix tests

* Fix coments

* Remove useless condition

* Fixed failed tests
 2020-03-17 17:31:01 +03:00
Alexander Alekhin
ca23c0e630 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-17 13:23:33 +03:00
Alexander Alekhin
683910f579 Merge pull request #16760 from dkurt:dnn_mobilenet_v3 2020-03-14 13:38:06 +00:00
Liubov Batanina
2645ee90ca
Merge pull request #16735 from l-bat:flatten_const_onnx 
* Supported Flatten for constant nodes

* Added default axis

* Refactoring

* Refactoring

* Added cast layer

* Fix comments

* Add Cast for layers
 2020-03-14 11:05:49 +00:00
Alexander Alekhin
d00e58cdb0 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-10 22:49:51 +00:00
Yashas Samaga B L
490908f0ff
Merge pull request #16436 from YashasSamaga:feature-enetb0-yolo 
dnn(darknet-importer): add grouped convolutions, sigmoid, swish, scale_channels

* update darknet importer to support enetb0-yolo

* remove dropout (pr16438) and fix formatting

* add test for scale_channels

* disable batch testing for scale channels

* do not set LayerParams::name

* merge all activations into setActivation
 2020-03-10 12:45:19 +03:00
Dmitry Kurtaev
b927ce18b2 Support for MobileNetV3-SSD from TensorFlow 2020-03-08 21:09:21 +03:00
Alexander Alekhin
619180dffd Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-06 20:41:30 +00:00
Alexander Alekhin
a76c72acdd Merge pull request #16738 from dkurt:onnx_broadcast 2020-03-06 19:04:43 +00:00
Alexander Alekhin
6271192a32 Merge pull request #16694 from alalek:dnn_disable_nn_builder_api 2020-03-06 13:18:41 +00:00
Alexander Alekhin
094a5b0cba Merge pull request #16719 from alalek:dnn_disable_nn_builder_api_3.4 2020-03-06 12:26:48 +00:00
Dmitry Kurtaev
9e332dc5fb Broadcasting from ONNX 2020-03-06 00:58:59 +03:00
Alexander Alekhin
d4a17da7b2 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-03-04 20:49:09 +00:00
Liubov Batanina
9ed1332355
Merge pull request #16722 from l-bat:reshape_opset_11 
* Supported Div op for constants

* Added Mul test
 2020-03-04 11:27:10 +03:00
ashishiva3@gmail.com
e18d5e94c7 Gather-Cast, Mul-Cast fusion 2020-03-03 21:08:28 +05:30
Alexander Alekhin
124bf8339f dnn(IE): use HAVE_DNN_IE_NN_BUILDER_2019 for NN Builder API code 
- CMake option: OPENCV_DNN_IE_NN_BUILDER_2019
 2020-03-03 08:07:54 +00:00
Alexander Alekhin
29d214474f dnn(IE): use HAVE_DNN_IE_NN_BUILDER_2019 for NN Builder API code 
- CMake option: OPENCV_DNN_IE_NN_BUILDER_2019
 2020-03-03 07:45:09 +00:00
Liubov Batanina
b1b78aedd2 Skipped ResizeUnfused test on Builder API 2020-03-02 15:45:29 +03:00
Alexander Alekhin
599a595c6f
Merge pull request #16573 from ashishkrshrivastava:opencvonnx 2020-02-29 21:20:40 +03:00
ashishiva3@gmail.com
8559237d4e ONNX: upsample subgraph fusion added 2020-02-29 15:24:06 +05:30
Alexander Alekhin
9c4b1d0a51 Merge pull request #16685 from YashasSamaga:cuda4dnn-fix-tests-pr16595 2020-02-27 20:42:45 +00:00
YashasSamaga
98b5fb4b9f disable MaskRCNN test for CUDA_FP16 2020-02-28 15:44:37 +05:30
Alexander Alekhin
cc0066d38d dnn(test): adjust check tolerance in Keypoints_face 2020-02-27 19:44:51 +03:00
Alexander Alekhin
45d073f889 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-02-26 20:09:03 +03:00
Dmitry Kurtaev
d8dea7896b
Merge pull request #16628 from dkurt:dnn_ngraph_custom_layers 
* Custom layers with nGraph

* nGraph: multiple outputs from nodes
 2020-02-26 17:51:18 +03:00
Alexander Alekhin
f48c84eaee Merge pull request #16656 from alalek:issue_16655 2020-02-26 12:47:46 +00:00
Alexander Alekhin
c2f5f5a202 dnn(test): configure filtering for 32-bit systems (part 2) 2020-02-24 18:18:33 +00:00
Alexander Alekhin
fce6e7248e Merge pull request #16629 from alalek:dnn_test_32bit 2020-02-23 17:09:31 +00:00
Alexander Alekhin
d54d01ca46 core(MatExpr): fix .type() bug 2020-02-23 17:05:05 +00:00
Alexander Alekhin
1540ae340f dnn(test): configure filtering for 32-bit systems 2020-02-22 17:35:26 +00:00
Alexander Alekhin
01048e5603
Merge pull request #16616 from alalek:dnn_fix_input_shape 
* dnn: fix processing of input shapes

- importer: avoid using of .setInput() => .setInputShape()
- setInput: shape limitation check (partial)

* dnn(test): test .setInput() in readNet()
 2020-02-21 22:39:54 +03:00
Alexander Alekhin
102ef39a27
Merge pull request #16225 from berak:dnn_pose_estimation 
dnn: small fix in pose estimation model
 2020-02-21 22:37:23 +03:00
Alexander Alekhin
96b26dc8f4 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-02-20 19:47:27 +03:00
berak
4fdf4d380c dnn: small fix in pose estimation model 2020-02-20 17:08:55 +01:00
Alexander Alekhin
30331eef4b Merge pull request #16595 from dkurt:dnn_ie_mask_rcnn 2020-02-20 12:50:27 +00:00
Dmitry Kurtaev
f3eef792eb Enable Mask R-CNN with Inference Engine. Full coverage with nGraph 2020-02-20 10:13:40 +03:00
Alexander Alekhin
1602a38fa9
Merge pull request #16572 from alalek:dnn_test_dldt_ir_v10 
* dnn(test): test DLDT IRv10 model, drop old models

* dnn(test): don't expect bitexact results in DLDT IR tests
 2020-02-18 17:57:46 +03:00
Liubov Batanina
e970eccbf1
Merge pull request #16472 from l-bat:cp_vton 
Add CP-VTON sample

* Support resize from PyTorch

* Add CP-VTON sample

* Fix downsampling

* Fix test

* Add model links

* Add default args

* Speed up resize

* Fix TOM link

* Add default args

* Fix comments

* Set aspect ratio for input

* Update links

* Check files exist
 2020-02-17 22:29:37 +03:00
Gagandeep Singh
a6f3a21256
Merge pull request #16424 from czgdp1807:issue-16370 
* fixed Split layer in ONNXImporter

* added test for fix of split layer

* fixed tests for Split layer

* applied reviews

* updated tests

* fixed paths in tests
 2020-02-15 14:12:20 +03:00
Alexander Alekhin
aa2777ed61 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-02-10 19:40:29 +03:00
Alexander Alekhin
db9f4436e0
Merge pull request #16473 from ashishkrshrivastava:opencvtf 2020-02-09 11:52:24 +03:00
ashishiva3@gmail.com
cd53144ed3 commit: Corrected end tensor_content parsing for StridedSlice layer. 2020-02-09 06:43:23 +05:30
Alexander Alekhin
0a691f7840 Merge pull request #16529 from alalek:dnn_skip_failed_ngraph_tests 2020-02-08 18:53:10 +00:00
Alexander Alekhin
8ecfb59930 dnn(test): skip failed ngraph tests 2020-02-07 22:43:40 +00:00
Alexander Alekhin
bcc9946650
Merge pull request #16438 from ashishkrshrivastava:opencvc 2020-02-07 23:17:59 +03:00
ashishiva3@gmail.com
d64529f6de Added parsing for Connected and Dropout. Modified maxpool padding 2020-02-07 07:26:36 +05:30
Alexander Alekhin
225566da7b Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-02-04 19:49:24 +03:00
Dmitry Kurtaev
005f38fb45 Fix dnn::ResizeLayer to manage varying input shapes 2020-02-04 09:06:17 +03:00
Alexander Alekhin
2ced568d34 Merge pull request #16220 from YashasSamaga:cuda4dnn-roi-pooling-test_fix-optim 2020-01-29 20:57:15 +00:00
Alexander Alekhin
560f85f8e5 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-01-28 14:26:57 +03:00
Alexander Alekhin
5429b1f5ff Merge pull request #16223 from l-bat:lip_jppnet 2020-01-27 19:17:43 +00:00
Alexander Alekhin
3d14dd4e39 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-01-22 16:58:30 +03:00
Alexander Alekhin
b21a861c5b dnn(test): adjust test tolerance for MYRIAD 2020-01-22 14:42:52 +03:00
Alexander Alekhin
16d8e9e0c8 Merge pull request #16358 from ashishkrshrivastava:opencv-2 2020-01-22 09:57:41 +00:00
ashishiva3@gmail.com
f29bc22005 Convolutional padding parsing modified 2020-01-19 22:33:12 +05:30
Yashas Samaga B L
d85e67d3ec Merge pull request #16063 from YashasSamaga:cuda4dnn-shortcut-unequal 
support eltwise sum with different number of input channels in CUDA backend

* add shortcut primitive

* add offsets in shortcut kernel

* skip tests involving more than two inputs

* remove redundant modulus operation

* support multiple inputs

* remove whole file indentation

* skip acc in0 trunc test if weighted

* use shortcut iff channels are unequal
 2020-01-16 21:54:00 +03:00
YashasSamaga
fd369a5004 fix and optimize ROIPooling 2020-01-15 22:53:48 +05:30
Alexander Alekhin
4cb9faf6c9 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2020-01-14 17:04:22 +03:00
Alexander Alekhin
a67228cd73 Merge pull request #16291 from dkurt:dnn_onnx_graph_simplifier 2020-01-14 12:45:59 +00:00
Dmitry Kurtaev
c1c84d2fd1 ONNX graphs simplifier 2020-01-14 12:45:49 +03:00
Dmitry Kurtaev
8f1e36f7c1 Disable some tests for Myriad target of nGraph 
Add lightweight IE hardware targets checks

nGraph: Concat with paddings

Enable more nGraph tests

Restore FP32->FP16 for GPU plugin of IE

try to fix buildbot

Use lightweight IE targets check only starts from R4
 2020-01-13 15:35:47 +03:00
Liubov Batanina
e9e3af0aaa Add global pool by axis test 2020-01-10 11:32:48 +03:00
Yashas Samaga B L
1fac1421e5 Merge pull request #16010 from YashasSamaga:cuda4dnn-fp16-tests 
* enable tests for DNN_TARGET_CUDA_FP16

* disable deconvolution tests

* disable shortcut tests

* fix typos and some minor changes

* dnn(test): skip CUDA FP16 test too (run_pool_max)
 2019-12-20 16:36:32 +03:00
Alexander Alekhin
97b6068c46 dnn(test): don't require downloaded data 2019-12-19 19:31:59 +00:00
Alexander Alekhin
4c86fc13cb Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-12-19 15:09:05 +03:00
Alexander Alekhin
4342657762 Merge pull request #16034 from Quantizs:irLoadFromBuffer 2019-12-19 10:00:07 +00:00
antalzsiroscandid
aa80f754f4 dnn: reading IR models from buffer 2019-12-18 15:31:08 +01:00
Alexander Alekhin
2c0d9fa81f dnn(test): fix Test_Model.Keypoints* tests 2019-12-16 18:07:23 +03:00
Alexander Alekhin
ba7b0f4c54 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-12-15 11:23:46 +00:00
Xuanda Yang
3d60a9b96c Merge pull request #16156 from TH3CHARLie:3.4 
* Eltwise::DIV support in Halide backend

* fix typo

* remove div from generated test suite to pass CI, switching to manual test...

* ensure divisor not near to zero

* use randu

* dnn(test): update test data for Eltwise.Accuracy/DIV layer test
 2019-12-13 18:29:39 +03:00
Diego
5b0b59ecfb Merge pull request #15189 from dvd42:keypoints_module 
Keypoints module
 2019-12-13 18:00:06 +03:00
Alexander Alekhin
92b9888837 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-12-12 13:02:19 +03:00
Alexander Alekhin
5ee7abbe3c
Merge pull request #16088 from alalek:dnn_eltwise_layer_different_src_channels 
dnn(eltwise): fix handling of different number of channels

* dnn(test): reproducer for Eltwise layer issue from PR16063

* dnn(eltwise): rework support for inputs with different channels

* dnn(eltwise): get rid of finalize(), variableChannels

* dnn(eltwise): update input sorting by number of channels

- do not swap inputs if number of channels are same after truncation

* dnn(test): skip "shortcut" with batch size 2 on MYRIAD targets
 2019-12-11 20:16:58 +03:00
Alexander Alekhin
202ba124a5 Merge pull request #16087 from YashasSamaga:cuda4dnn-eltwise-div 2019-12-06 18:33:55 +00:00
YashasSamaga
a91eca6ec2 add DIV support to EltwiseOp 2019-12-06 21:28:36 +05:30
Alexander Alekhin
8108fb0575 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-12-05 18:27:45 +03:00
Alexander Alekhin
986c5084a4 Merge pull request #16036 from dkurt:dnn_backport_15203 2019-12-05 09:56:00 +00:00
Dmitry Kurtaev
d8e10f3a8d Enable MaxPooling with indices in Inference Engine 2019-12-04 19:14:55 +03:00
Alexander Alekhin
1633c29f29 Merge pull request #16037 from alalek:dnn_test_fix_skip_vulkan 2019-12-04 15:49:47 +00:00
Alexander Alekhin
0dc0bc80ae dnn(test): fix Vulkan skip test tag 2019-12-02 18:22:10 +03:00
Dmitry Kurtaev
ca1ba7a53d Backport for dnn input shape estimation 2019-12-02 16:28:59 +03:00
Alexander Alekhin
4b0132ed7a Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-12-02 16:26:52 +03:00
Lubov Batanina
7523c777c5 Merge pull request #15537 from l-bat:ngraph 
* Support nGraph

* Fix resize
 2019-12-02 16:16:06 +03:00
Manjunath Bhat
78c5e41c23 Merge pull request #15808 from thebhatman:Mish_swish 
* Added Swish and Mish activations

* Fixed whitespace errors

* Kernel implementation done

* Added function for launching kernel

* Changed type of 1.0

* Attempt to add test for Swish and Mish

* Resolving type mismatch for log

* exp from device

* Use log1pexp instead of adding 1

* Added openCL kernels
 2019-12-02 00:06:17 +03:00
thebhatman
8a18d132fc Port Swish and Mish layers 2019-12-01 11:55:39 +03:00
Alexander Alekhin
b6a58818bb Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-11-11 20:25:42 +00:00
Lubov Batanina
cfc781949d Merge pull request #15811 from l-bat:eltwise_div 
Supported ONNX Squeeze, ReduceL2 and Eltwise::DIV

* Support eltwise div

* Fix test

* OpenCL support added

* refactoring

* fix code style

* Only squeeze with axes supported
 2019-11-09 14:11:09 +03:00
Dimitri Gerin
7c4158d8c2 Fix dnn::getLayerInputs 2019-11-07 08:13:33 +03:00
Alexander Alekhin
055ffc0425 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-10-24 18:21:19 +00:00
Alexander Alekhin
d65fead337 Merge pull request #15752 from dkurt:fix_15750 2019-10-24 07:06:32 +00:00
Dmitry Kurtaev
dfe0368835 Fix custom IE layers in case of no MKLDNN plugin 2019-10-21 19:09:44 +03:00
Yashas Samaga B L
613c12e590 Merge pull request #14827 from YashasSamaga:cuda4dnn-csl-low 
CUDA backend for the DNN module

* stub cuda4dnn design

* minor fixes for tests and doxygen

* add csl public api directory to module headers

* add low-level CSL components

* add high-level CSL components

* integrate csl::Tensor into backbone code

* switch to CPU iff unsupported; otherwise, fail on error

* add fully connected layer

* add softmax layer

* add activation layers

* support arbitary rank TensorDescriptor

* pass input wrappers to `initCUDA()`

* add 1d/2d/3d-convolution

* add pooling layer

* reorganize and refactor code

* fixes for gcc, clang and doxygen; remove cxx14/17 code

* add blank_layer

* add LRN layer

* add rounding modes for pooling layer

* split tensor.hpp into tensor.hpp and tensor_ops.hpp

* add concat layer

* add scale layer

* add batch normalization layer

* split math.cu into activations.cu and math.hpp

* add eltwise layer

* add flatten layer

* add tensor transform api

* add asymmetric padding support for convolution layer

* add reshape layer

* fix rebase issues

* add permute layer

* add padding support for concat layer

* refactor and reorganize code

* add normalize layer

* optimize bias addition in scale layer

* add prior box layer

* fix and optimize normalize layer

* add asymmetric padding support for pooling layer

* add event API

* improve pooling performance for some padding scenarios

* avoid over-allocation of compute resources to kernels

* improve prior box performance

* enable layer fusion

* add const layer

* add resize layer

* add slice layer

* add padding layer

* add deconvolution layer

* fix channelwise  ReLU initialization

* add vector traits

* add vectorized versions of relu, clipped_relu, power

* add vectorized concat kernels

* improve concat_with_offsets performance

* vectorize scale and bias kernels

* add support for multi-billion element tensors

* vectorize prior box kernels

* fix address alignment check

* improve bias addition performance of conv/deconv/fc layers

* restructure code for supporting multiple targets

* add DNN_TARGET_CUDA_FP64

* add DNN_TARGET_FP16

* improve vectorization

* add region layer

* improve tensor API, add dynamic ranks

1. use ManagedPtr instead of a Tensor in backend wrapper
2. add new methods to tensor classes
  - size_range: computes the combined size of for a given axis range
  - tensor span/view can be constructed from a raw pointer and shape
3. the tensor classes can change their rank at runtime (previously rank was fixed at compile-time)
4. remove device code from tensor classes (as they are unused)
5. enforce strict conditions on tensor class APIs to improve debugging ability

* fix parametric relu activation

* add squeeze/unsqueeze tensor API

* add reorg layer

* optimize permute and enable 2d permute

* enable 1d and 2d slice

* add split layer

* add shuffle channel layer

* allow tensors of different ranks in reshape primitive

* patch SliceOp to allow Crop Layer

* allow extra shape inputs in reshape layer

* use `std::move_backward` instead of `std::move` for insert in resizable_static_array

* improve workspace management

* add spatial LRN

* add nms (cpu) to region layer

* add max pooling with argmax ( and a fix to limits.hpp)

* add max unpooling layer

* rename DNN_TARGET_CUDA_FP32 to DNN_TARGET_CUDA

* update supportBackend to be more rigorous

* remove stray include from preventing non-cuda build

* include op_cuda.hpp outside condition #if

* refactoring, fixes and many optimizations

* drop DNN_TARGET_CUDA_FP64

* fix gcc errors

* increase max. tensor rank limit to six

* add Interp layer

* drop custom layers; use BackendNode

* vectorize activation kernels

* fixes for gcc

* remove wrong assertion

* fix broken assertion in unpooling primitive

* fix build errors in non-CUDA build

* completely remove workspace from public API

* fix permute layer

* enable accuracy and perf. tests for DNN_TARGET_CUDA

* add asynchronous forward

* vectorize eltwise ops

* vectorize fill kernel

* fixes for gcc

* remove CSL headers from public API

* remove csl header source group from cmake

* update min. cudnn version in cmake

* add numerically stable FP32 log1pexp

* refactor code

* add FP16 specialization to cudnn based tensor addition

* vectorize scale1 and bias1 + minor refactoring

* fix doxygen build

* fix invalid alignment assertion

* clear backend wrappers before allocateLayers

* ignore memory lock failures

* do not allocate internal blobs

* integrate NVTX

* add numerically stable half precision log1pexp

* fix indentation, following coding style,  improve docs

* remove accidental modification of IE code

* Revert "add asynchronous forward"

This reverts commit 1154b9da9da07e9b52f8a81bdcea48cf31c56f70.

* [cmake] throw error for unsupported CC versions

* fix rebase issues

* add more docs, refactor code, fix bugs

* minor refactoring and fixes

* resolve warnings/errors from clang

* remove haveCUDA() checks from supportBackend()

* remove NVTX integration

* changes based on review comments

* avoid exception when no CUDA device is present

* add color code for CUDA in Net::dump
 2019-10-21 14:28:00 +03:00
Dmitry Kurtaev
af61a15839 Fix Darknet eltwise 2019-10-19 12:54:15 +03:00
Dmitry Kurtaev
adbd613660 Enable Eltwise layer with different numbers of inputs channels 2019-10-18 18:51:52 +03:00
Alexander Alekhin
626bfbf309 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-10-05 15:45:31 +00:00
Alexander Alekhin
feff8bf972 Merge pull request #15626 from alalek:dnn_openvino_2019r3 2019-10-04 19:45:37 +00:00
Alexander Alekhin
c13a5ce229 Merge pull request #15622 from dkurt:enet_ie_cpu 2019-10-04 16:31:05 +00:00
Dmitry Kurtaev
e35fd463e7 Enable ENet with Inference Engine backend on CPU 2019-10-04 18:10:11 +03:00
Alexander Alekhin
fd11e3a81d dnn: update IE tests 2019-10-04 17:49:10 +03:00
Alexander Alekhin
f0058bbed3 dnn(test): fix optional test data 2019-10-04 17:26:35 +03:00
Alexander Alekhin
3fb6617d62 Merge remote-tracking branch 'upstream/3.4' into merge-3.4 2019-10-02 17:49:19 +03:00
Alexander Alekhin
c06115cb3f Merge pull request #15619 from alalek:dnn_eltwise_sum_ie_ocl 2019-10-01 18:02:45 +00:00
Alexander Alekhin
8814645c8d dnn(test): skip IE/OCL test for "sum" 2019-10-01 18:29:47 +03:00