platform/upstream/dldt.git
3 years ago[IE][VPU]: Mish decomposition (#3412)
Maksim Doronin [Mon, 30 Nov 2020 15:51:51 +0000 (18:51 +0300)]
[IE][VPU]: Mish decomposition (#3412)

Introduce vpu ngraph transformation that decomposes Mish layer with the following formula: mish(x) = x * tanh(ln(exp(x) + 1)). It is needed to fix the accuracy drop of Yolo-v4 public model that happened due to Mish implementation in the myriad plugin is not accurate enough. It is just a workaround and will be removed when Mish layer will be improved.
map: 70.97% (ref: 70.96%)
coco_precision: 73.50% (ref 73.49%)

3 years agoFix paths for `squeezenet1.1` in time_tests config (#3417)
Vitaliy Urusovskij [Mon, 30 Nov 2020 15:45:15 +0000 (18:45 +0300)]
Fix paths for `squeezenet1.1` in time_tests config (#3417)

3 years ago[DOC] ShapeInference.md update. slyalin comments (#3355)
Evgenya Stepyreva [Mon, 30 Nov 2020 13:39:58 +0000 (16:39 +0300)]
[DOC] ShapeInference.md update. slyalin comments (#3355)

* [DOC] ShapeInference.md update. slyalin comments

* Apply suggestions from code review

Co-authored-by: Alina Alborova <alina.alborova@intel.com>
* Apply suggestions from code review

Co-authored-by: Alina Alborova <alina.alborova@intel.com>
* Update docs/IE_DG/ShapeInference.md

Co-authored-by: Alina Alborova <alina.alborova@intel.com>
3 years agoUpdated transformations API guide (#3386)
Gleb Kazantaev [Mon, 30 Nov 2020 10:50:30 +0000 (13:50 +0300)]
Updated transformations API guide (#3386)

* Updated transformations docs

* Added RTTI to examples

* Fixed grammar mistakes

* Mention Sinks in NG docs

3 years ago[install_prerequisites.sh] upgrade pip to install tensorflow (#3392) (#3402)
Alexey Suhov [Fri, 27 Nov 2020 15:35:28 +0000 (18:35 +0300)]
[install_prerequisites.sh] upgrade pip to install tensorflow (#3392) (#3402)

3 years ago[GNA] Fix import model with header version 2.1 (#3363)
Andrey Dmitriev [Fri, 27 Nov 2020 14:01:59 +0000 (17:01 +0300)]
[GNA] Fix import model with header version 2.1 (#3363)

3 years ago[IE][VPU]: Use unified pcie firmware (#3383)
Maksim Doronin [Fri, 27 Nov 2020 12:25:37 +0000 (15:25 +0300)]
[IE][VPU]: Use unified pcie firmware (#3383)

* Unified firmware named pcie-ma2x8x is needed to use both ma2480 and ma2085 with pcie connection.
* Update firmware built from R14 release branch

3 years ago[IE CLDNN] Fix handling of fsv4 input padding in mmad bfyx to fsv32 convolution kerne...
Jedrzej Hajduczenia [Fri, 27 Nov 2020 10:41:13 +0000 (11:41 +0100)]
[IE CLDNN] Fix handling of fsv4 input padding in mmad bfyx to fsv32 convolution kernel (#3253)

3 years ago[CPU] NMS test for the rest of 1st & 2nd outputs. (#3380)
Nikolay Shchegolev [Fri, 27 Nov 2020 10:21:52 +0000 (13:21 +0300)]
[CPU] NMS test for the rest of 1st & 2nd outputs. (#3380)

3 years agoCherry-pick from master (#3379)
Alexander Peskov [Thu, 26 Nov 2020 16:37:00 +0000 (19:37 +0300)]
Cherry-pick from master  (#3379)

* [DOC] Add macOS specific section into intro guide

Signed-off-by: Peskov <alexander.peskov@intel.com>
* Remove libinference_engine_legacy from 'must link with' section

Signed-off-by: Peskov <alexander.peskov@intel.com>
3 years ago[IE][VPU]: KW fixes 2021.2 (#3313)
Maksim Doronin [Thu, 26 Nov 2020 14:00:29 +0000 (17:00 +0300)]
[IE][VPU]: KW fixes 2021.2 (#3313)

3 years agoFixes for documentation build (#3365)
Ilya Lavrenov [Thu, 26 Nov 2020 11:32:33 +0000 (14:32 +0300)]
Fixes for documentation build (#3365)

3 years ago[IE][VPU]: Introduces ENABLE_EARLY_ELTWISE_RELU_FUSION option (#3343)
Gladilov, Gleb [Thu, 26 Nov 2020 10:27:53 +0000 (13:27 +0300)]
[IE][VPU]: Introduces ENABLE_EARLY_ELTWISE_RELU_FUSION option (#3343)

Unconditional Eltwise+ReLU merge makes some models take a bit more memory,
so they could not be inferred on Myriad-X. With this change used can
optionally disable this optimization and use Myriad-X.

3 years agoCherry-pick #3295 to releases/2021/2 (#3353)
Andrey Zaytsev [Wed, 25 Nov 2020 15:08:22 +0000 (18:08 +0300)]
Cherry-pick #3295 to releases/2021/2 (#3353)

* Feature/azaytsev/change layout (#3295)

* Updated openvino_docs.xml

3 years agoFix the potential unlimited mem allocation problem (#3347)
Tomasz Dołbniak [Wed, 25 Nov 2020 14:41:25 +0000 (15:41 +0100)]
Fix the potential unlimited mem allocation problem (#3347)

3 years agoAdd WA for OMZ Keras models in stress tests (#3341)
Vitaliy Urusovskij [Wed, 25 Nov 2020 11:38:15 +0000 (14:38 +0300)]
Add WA for OMZ Keras models in stress tests (#3341)

3 years agoAdd `fillBlobs` to fix issue with negative values in `image_info` input layer (#3335)
Vitaliy Urusovskij [Wed, 25 Nov 2020 11:37:21 +0000 (14:37 +0300)]
Add `fillBlobs` to fix issue with negative values in `image_info` input layer (#3335)

3 years agoFixes for NMS documentation. (#3305)
Vladimir Gavrilov [Wed, 25 Nov 2020 11:05:14 +0000 (14:05 +0300)]
Fixes for NMS documentation. (#3305)

* Fixes for NMS documentation.

* Small fixes.

* Small fix.

3 years agoFixed klockwork for legacy library (#3320)
Ilya Lavrenov [Wed, 25 Nov 2020 08:14:55 +0000 (11:14 +0300)]
Fixed klockwork for legacy library (#3320)

3 years ago[ DOC ] setBatchSize update (#3316)
Evgenya Stepyreva [Tue, 24 Nov 2020 17:56:27 +0000 (20:56 +0300)]
[ DOC ] setBatchSize update (#3316)

* [ DOC ] setBatchSize update

* comments

* Links fixed

* Update docs/IE_DG/ShapeInference.md

* Fixing broken note

* Fixing broken note [2]

* Update docs/IE_DG/ShapeInference.md

Co-authored-by: Alina Alborova <alina.alborova@intel.com>
Co-authored-by: Alina Alborova <alina.alborova@intel.com>
3 years ago[Scripts] install_NEO_OCL_driver: Updated capturing '-y' option. Removed installing...
Artyom Anokhov [Tue, 24 Nov 2020 17:20:21 +0000 (20:20 +0300)]
[Scripts] install_NEO_OCL_driver: Updated capturing '-y' option. Removed installing prerequisites. (#3322)

* install_NEO_OCL_driver: Updated capturing '-y' option. Removed installing prerequisites.

* install_NEO_OCL_driver: Rollback removing install_prerequisites, but keep only ocl-icd & libnuma dependencies. Updated definition of UBUNTU_VERSION.

3 years ago[CPU] fixed MergePermuteAndReorder optimization (#3317)
Anton Voronov [Tue, 24 Nov 2020 16:14:30 +0000 (19:14 +0300)]
[CPU] fixed MergePermuteAndReorder optimization (#3317)

3 years agoFixed KW issues reported for build 754 of speech recognition demo (#3326)
Piotr Rozen [Tue, 24 Nov 2020 15:15:45 +0000 (16:15 +0100)]
Fixed KW issues reported for build 754 of speech recognition demo (#3326)

3 years ago[GNA] Fix a global buffer overflow in GNAModelSerial::Import (#3290) (#3327)
Elizaveta Lobanova [Tue, 24 Nov 2020 15:12:16 +0000 (18:12 +0300)]
[GNA] Fix a global buffer overflow in GNAModelSerial::Import (#3290) (#3327)

3 years ago[IE][VPU]: Update call signature with the fw from R14 develop branch (#3241)
Krishna Prabhakaran [Tue, 24 Nov 2020 13:22:56 +0000 (05:22 -0800)]
[IE][VPU]: Update call signature with the fw from R14 develop branch (#3241)

* Update call signature with the firmware from mdk R14 branch
* Update VPU firmware build number

3 years agoFixed links to build instruction (#3309)
Ilya Lavrenov [Tue, 24 Nov 2020 10:12:22 +0000 (13:12 +0300)]
Fixed links to build instruction (#3309)

3 years agoadded state API changes to doc (#3307)
Svetlana Dolinina [Tue, 24 Nov 2020 09:17:37 +0000 (12:17 +0300)]
added state API changes to doc (#3307)

3 years agoAlign time_tests models with master (#3270)
Vitaliy Urusovskij [Tue, 24 Nov 2020 08:06:37 +0000 (11:06 +0300)]
Align time_tests models with master (#3270)

* Add new model to `tgl_test_config.yml` (#3236)

* Fix wrong path for `yolo-v2-tiny-ava-0001` for time_tests

* Add several new models to `tgl_test_config.yml` in time_tests

3 years ago[Scripts] Update install_NEO_OCL_driver with 20.35.17767 for 2021.2 (#3288)
Artyom Anokhov [Mon, 23 Nov 2020 18:11:27 +0000 (21:11 +0300)]
[Scripts] Update install_NEO_OCL_driver with 20.35.17767 for 2021.2 (#3288)

* install_NEO_OCL_driver: Added new CLI options. Added help. Added new available to download and install driver - 20.35.17767

* install_NEO_OCL_driver: Fixed descriptions and var names

* install_NEO_OCL_driver: Updated syntax

3 years agoFixed static analysis issues (transformations) (#3276)
Maxim Andronov [Mon, 23 Nov 2020 14:08:54 +0000 (17:08 +0300)]
Fixed static analysis issues (transformations) (#3276)

3 years ago[LPT] integration: issue #42391 & issue #43001 (#3201)
Edward Shogulin [Mon, 23 Nov 2020 14:08:41 +0000 (17:08 +0300)]
[LPT] integration: issue #42391 & issue #43001  (#3201)

* [LPT] NetworkHelper::roundWithTolerance: removed tolerance & rename to round
[LPT] NetworkHelper::round functional tests
[LPT] ieFuncTests: updated some test-cases

* [LPT] Subtract is not used

* [LPT] AddTransformation: zero handling

* [LPT] AddTransformation test

3 years ago[IE CLDNN] Gather fix (#3282)
Sergey Shlyapnikov [Mon, 23 Nov 2020 13:40:17 +0000 (16:40 +0300)]
[IE CLDNN] Gather fix (#3282)

3 years agoFixed static analysis issues (#3254)
Maxim Andronov [Sun, 22 Nov 2020 17:42:53 +0000 (20:42 +0300)]
Fixed static analysis issues (#3254)

3 years ago[CPU] NormalizeL2. Support 3D layout. (#3256)
Nikolay Shchegolev [Sun, 22 Nov 2020 17:18:33 +0000 (20:18 +0300)]
[CPU] NormalizeL2. Support 3D layout. (#3256)

3 years ago[CPU] NMS. Fill the rest of first out blob. (#3251)
Nikolay Shchegolev [Sun, 22 Nov 2020 17:06:15 +0000 (20:06 +0300)]
[CPU] NMS. Fill the rest of first out blob. (#3251)

3 years agofix for sliced input, fix for reshaping of Loop op (squash commits) (#3261)
Ivan Tikhonov [Sun, 22 Nov 2020 10:25:04 +0000 (13:25 +0300)]
fix for sliced input, fix for reshaping of Loop op (squash commits) (#3261)

3 years ago[ MO ] Fix for preprocessing insertion on ShapeOf paths (#3233)
Evgenya Stepyreva [Fri, 20 Nov 2020 13:25:11 +0000 (16:25 +0300)]
[ MO ] Fix for preprocessing insertion on ShapeOf paths (#3233)

* [ MO ] Fix for preprocessing insertion on ShapeOf paths

* ScaleInput tests

3 years ago[IE CLDNN] Minor fixed to NMS and TopK (#3224)
Vladimir Paramuzov [Fri, 20 Nov 2020 12:37:32 +0000 (15:37 +0300)]
[IE CLDNN] Minor fixed to NMS and TopK (#3224)

3 years ago[ nG: TI ] Fix for reshaping TensorIterator (#3237)
Evgenya Stepyreva [Fri, 20 Nov 2020 12:22:19 +0000 (15:22 +0300)]
[ nG: TI ] Fix for reshaping TensorIterator (#3237)

* [ nG: TI ] Fix for reshaping TensorIterator

* Trigger CI

* misprint

* style

* style

3 years agoFix "results" uploading to DB in time_tests (#3214)
Vitaliy Urusovskij [Fri, 20 Nov 2020 10:25:37 +0000 (13:25 +0300)]
Fix "results" uploading to DB in time_tests (#3214)

3 years agoComplete CTCGreedyDecoder implementation in nGraph and add tests (#3206)
Roman Kazantsev [Fri, 20 Nov 2020 08:05:45 +0000 (11:05 +0300)]
Complete CTCGreedyDecoder implementation in nGraph and add tests (#3206)

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
3 years agoUpdate tf2 yolov4 docs support (#3227)
iliya mironov [Fri, 20 Nov 2020 06:24:14 +0000 (09:24 +0300)]
Update tf2 yolov4 docs support (#3227)

Co-authored-by: Your Name <you@example.com>
3 years ago[ MO TF ] TF FakeQuantize normalization fix (#3180)
Evgenya Stepyreva [Thu, 19 Nov 2020 15:38:20 +0000 (18:38 +0300)]
[ MO TF ] TF FakeQuantize normalization fix (#3180)

3 years agoUpdate install_openvino_dependencies.sh (#3196)
Dmitrii Denisov [Thu, 19 Nov 2020 14:38:44 +0000 (17:38 +0300)]
Update install_openvino_dependencies.sh (#3196)

3 years ago[LPT] Transpose, MatMul: Klocwork fixes (#3192)
Vladislav Golubev [Thu, 19 Nov 2020 12:15:39 +0000 (15:15 +0300)]
[LPT] Transpose, MatMul: Klocwork fixes (#3192)

3 years agoFix NMS1/3/4 to Legacy conversion (#3210)
Gleb Kazantaev [Thu, 19 Nov 2020 11:20:32 +0000 (14:20 +0300)]
Fix NMS1/3/4 to Legacy conversion (#3210)

* Fix NMS1/3/4 to Legacy conversion

* Added functoinal tests

3 years ago[IE] - Interpolate ONNX workaround (#3203)
Andrey Sokolov [Thu, 19 Nov 2020 08:59:46 +0000 (11:59 +0300)]
[IE] - Interpolate ONNX workaround (#3203)

Co-authored-by: DariaMityagina <daria.mityagina@intel.com>
3 years agoBackport of PR #3092 - TopK sorting fix (#3195)
Tomasz Jankowski [Thu, 19 Nov 2020 08:40:45 +0000 (09:40 +0100)]
Backport of PR #3092 - TopK sorting fix (#3195)

3 years agoDisable v10 serealizer (#3184)
Ilya Churaev [Thu, 19 Nov 2020 08:34:59 +0000 (11:34 +0300)]
Disable v10 serealizer (#3184)

* Disable v10 serealizer

* Fixed comments

3 years agoTemporaly skip GRU Cell test (#3183)
Ivan Tikhonov [Tue, 17 Nov 2020 18:27:25 +0000 (21:27 +0300)]
Temporaly skip GRU Cell test (#3183)

3 years agoFix accuracy problem for YoloV3 (incorrect StridedSlice nop elimination) (#3179)
Mateusz Bencer [Tue, 17 Nov 2020 17:12:01 +0000 (18:12 +0100)]
Fix accuracy problem for YoloV3 (incorrect StridedSlice nop elimination) (#3179)

3 years ago[IE][VPU]: Align buffer conditions with the device part (#3168)
Andrey Sokolov [Tue, 17 Nov 2020 16:40:23 +0000 (19:40 +0300)]
[IE][VPU]: Align buffer conditions with the device part (#3168)

3 years ago[GNA] Security fixes for 2021.2 (#3167)
Elizaveta Lobanova [Tue, 17 Nov 2020 16:26:17 +0000 (19:26 +0300)]
[GNA] Security fixes for 2021.2 (#3167)

3 years agoONNX LSTM/GRU/RNN seqence_lens input tests with reverse and bidirectional mode (...
Katarzyna Mitrus [Tue, 17 Nov 2020 15:44:03 +0000 (16:44 +0100)]
ONNX LSTM/GRU/RNN seqence_lens input tests with reverse and bidirectional mode (#3149)

3 years agoFix NMS5 dynamism issues (#3137)
Gleb Kazantaev [Tue, 17 Nov 2020 15:34:05 +0000 (18:34 +0300)]
Fix NMS5 dynamism issues (#3137)

* Fixed ConvertNMS5ToLegacy conversion pass to resolve dynamism as soon as possible; fixed output precision to avoid troubles during convert_precision pass

* Added NMS5ToLegacy conversion for cldnn plugin

* Fixed NMS3 copy method

* Updated NMS5ToLegacy pass to avoid convert precision issues

* Updated nGraph reader and Functional tests

* Removing odd CF call

* removed xfail for test that is expected to pass

* Fixed Convert operations friendly_name

* Removed NMS5ToLegacy from opset1_to_legacy conversion

* Update NMS5 layer tests

* Update NMS5 DSR tests

3 years ago[LPT] implementation were moved from common dir (#3173)
Edward Shogulin [Tue, 17 Nov 2020 15:06:50 +0000 (18:06 +0300)]
[LPT] implementation were moved from common dir (#3173)

3 years ago[CPU] Normalize node. Broadcast weights from 1 to channels. (#3141)
Nikolay Shchegolev [Tue, 17 Nov 2020 14:01:44 +0000 (17:01 +0300)]
[CPU] Normalize node. Broadcast weights from 1 to channels. (#3141)

3 years agoAdded SECURITY.md back (#3177)
Ilya Lavrenov [Tue, 17 Nov 2020 13:44:44 +0000 (16:44 +0300)]
Added SECURITY.md back (#3177)

3 years ago[DOC] Update Docker install guide (#3055)
Kate Generalova [Tue, 17 Nov 2020 13:43:56 +0000 (16:43 +0300)]
[DOC] Update Docker install guide (#3055)

* [DOC] Update Docker install guide

* [DOC] Add proxy for Windows Docker install guide

* [DOC] move up prebuilt images section

* Update installing-openvino-linux.md

* Update installing-openvino-docker-linux.md

* Update installing-openvino-docker-linux.md

Formatting fixes

* Update installing-openvino-docker-linux.md

Fixed formatting issues

* Update installing-openvino-docker-windows.md

Minor fixes

* Update installing-openvino-docker-linux.md

Fixed formatting issues

* [DOC] update text with CPU image, remove proxy for win

* Update installing-openvino-docker-windows.md

Minor fixes

* Update installing-openvino-docker-windows.md

Minor fix

* Update installing-openvino-docker-windows.md

Minor fix

* Update installing-openvino-docker-windows.md

Co-authored-by: Andrey Zaytsev <andrey.zaytsev@intel.com>
3 years ago[IE][VPU]: Refactoring of SpecialStageProcessor (#2885)
Roman Vyunov (Intel) [Tue, 17 Nov 2020 13:30:30 +0000 (16:30 +0300)]
[IE][VPU]: Refactoring of SpecialStageProcessor (#2885)

* SpecialStageProcessor refactoring
* Fix for Yolo-v3-pytorch and related test

3 years agoFix graph clenup (#3159)
iliya mironov [Tue, 17 Nov 2020 13:28:27 +0000 (16:28 +0300)]
Fix graph clenup (#3159)

* Fix graph clenup

* Refactoring graph clean up function

* Change wa comment

Co-authored-by: Your Name <you@example.com>
3 years agoRevert "Reduced usage of batch in python samples (#3104)" (#3171)
Mikhail Ryzhov [Tue, 17 Nov 2020 11:57:40 +0000 (14:57 +0300)]
Revert "Reduced usage of batch in python samples (#3104)" (#3171)

This reverts commit 23e653858bc5e543afe17dcfdb6ce89047e96788.

3 years ago[IE CLDNN] program_impl::replace_all_usages causes invalid memory access. (#138)
chxd [Tue, 17 Nov 2020 10:45:42 +0000 (18:45 +0800)]
[IE CLDNN] program_impl::replace_all_usages causes invalid memory access. (#138)

3 years agoadded ubuntu 20.04 support to speech recognition demo (#3145)
Piotr Rozen [Tue, 17 Nov 2020 10:32:14 +0000 (11:32 +0100)]
added ubuntu 20.04 support to speech recognition demo (#3145)

updated speech recognition demo resources URLs to AWS

3 years agoPrint GNA Lib function name when error code obtained (#3129)
Krzysztof Bruniecki [Tue, 17 Nov 2020 10:16:15 +0000 (11:16 +0100)]
Print GNA Lib function name when error code obtained (#3129)

3 years agoTI and Sequences related transformations: transition from cnn to ngraph versions...
Ivan Tikhonov [Tue, 17 Nov 2020 09:38:22 +0000 (12:38 +0300)]
TI and Sequences related transformations: transition from cnn to ngraph versions (#2927)

* sequences to ti transformations, support for seq_lengths input, update reference implemetations, add new tests

* fix python api, update sequences to ti transformation

* transition from cnn to ngraph transformations for cpu, gpu, vpu plugins

* fix convert_ti_to_sequence transformation

* fix naming issue in unroll transformation

* test pure TensorIterator in vpu plugin

* fix sequences to ti transformation

* Update sequences to TI transformation: fix reverse sequence support

* update single layer tests, fix TI reference impl, fix Sequences to TI transformations

* ngraph code style

* fix build

* fix ngraph python api

* resolver review comments, refactoring

* revert vpu changes

* disable/fix tests

* refactoring

* Resolve review remarks

* optimization of LSTMSeq -> LSTMSeq IE: remove unnecessary Transpose ops

* Refactoring of transformation pipeline in cpu and gpu plugins, align GRU/RNN -> GRU/RNN IE with LSTM -> LSTM IE

* update TensorIterator tests, refactoring

* fix typo

* Fix unit tests, delete unnecessary callbacks

* Refactoring: delete commented code

* Add FullyConnected to skipConstInfer list for legacy ConstFolding

* disable legacy cnn unit tests

* delete xfail

* fix for backward compatibility with opset1::LSTMCell

* delete xfail

* fix build, remove Reshape layer from skipConstInfer list

3 years ago[nGraph][ONNX] Rename xfails (#3061)
Jan Iwaszkiewicz [Tue, 17 Nov 2020 09:34:33 +0000 (10:34 +0100)]
[nGraph][ONNX] Rename xfails (#3061)

3 years agoExtend information to submit to a DB in time_tests (#3018)
Vitaliy Urusovskij [Tue, 17 Nov 2020 09:07:40 +0000 (12:07 +0300)]
Extend information to submit to a DB in time_tests (#3018)

* Optimize imports in time_tests conftest.py

* Extend information to submit to a DB in time_tests:
1. Add `--manifest` and `--db_metadata` CLI keys
2. Add `prepare_db_info` fixture

* Move `validate_test_case` fixture to a const schema, remove extra checks

* Add `manifest_metadata` fixture to parse and validate manifest only once

* Add OS utils to submit os info to DB

3 years agoFix ApplyPermutations for the case when more then 1 input with permutes (#3095)
Maxim Vafin [Tue, 17 Nov 2020 08:58:01 +0000 (11:58 +0300)]
Fix ApplyPermutations for the case when more then 1 input with permutes (#3095)

3 years agoRemoved documents which are ported to OpenVINO WiKi (#3106)
Ilya Lavrenov [Tue, 17 Nov 2020 08:46:05 +0000 (11:46 +0300)]
Removed documents which are ported to OpenVINO WiKi (#3106)

* Removed documents which are ported to OpenVINO WiKi

* More updates

* Removed getting started

* Ported more guides to OpenVINO Wiki

3 years agoConnect some ngraph and IE cmake options (#3147)
Ilya Lavrenov [Tue, 17 Nov 2020 08:42:34 +0000 (11:42 +0300)]
Connect some ngraph and IE cmake options (#3147)

3 years ago[IE][TOOLS] Fix for CCT two IR mode (#3084)
Roman Vyunov (Intel) [Tue, 17 Nov 2020 08:41:48 +0000 (11:41 +0300)]
[IE][TOOLS] Fix for CCT two IR mode (#3084)

3 years ago[IE][VPU]: Fix condition in HW tiling (#3111)
Aleksandr Korolev [Tue, 17 Nov 2020 08:36:23 +0000 (11:36 +0300)]
[IE][VPU]: Fix condition in HW tiling (#3111)

Co-authored-by: kora6 <kora6@github.com>
3 years ago[ngraph] Fix compare functions for functional tests (#3161)
Aleksandr Pertovsky [Tue, 17 Nov 2020 08:32:52 +0000 (11:32 +0300)]
[ngraph] Fix compare functions for functional tests (#3161)

3 years ago[IE CLDNN] Fixed mem leak (#3153)
Vladimir Paramuzov [Tue, 17 Nov 2020 08:18:48 +0000 (11:18 +0300)]
[IE CLDNN] Fixed mem leak (#3153)

3 years agoFix compilation error in gna_wait_test (#3150)
Mateusz Tabaka [Tue, 17 Nov 2020 08:08:01 +0000 (09:08 +0100)]
Fix compilation error in gna_wait_test (#3150)

error: converting to ‘std::vector<std::tuple<unsigned int, long int,......

3 years agoShow help message when -i or -m is not set (#3144)
Anton Romanov [Tue, 17 Nov 2020 07:50:23 +0000 (10:50 +0300)]
Show help message when -i or -m is not set (#3144)

3 years ago[CPU] Interpolate operation improvements (#2366)
Chenhu Wang [Tue, 17 Nov 2020 07:42:34 +0000 (15:42 +0800)]
[CPU] Interpolate operation improvements (#2366)

* interpolate improvement

* JITTED cubic mode

* fix 'code is too big' when JIT

* extend test to cover tail code path

* transformation of interpolate1 to interpolate4

* add low precision transformation for interpolate4

3 years agoReduced usage of batch in python samples (#3104)
Mikhail Ryzhov [Tue, 17 Nov 2020 07:12:33 +0000 (10:12 +0300)]
Reduced usage of batch in python samples (#3104)

* Reduced usage of batch in python sampes

Excluded from hello_classification and object_detection samples

3 years ago[IE TESTS] Move main.cpp to shared lib from plugin functional tests (#3157)
Irina Efode [Tue, 17 Nov 2020 07:11:17 +0000 (10:11 +0300)]
[IE TESTS] Move main.cpp to shared lib from plugin functional tests (#3157)

3 years ago[IE TESTS][IE CPU] Add smoke prefix for broadcast SLT (#3158)
Irina Efode [Tue, 17 Nov 2020 07:10:45 +0000 (10:10 +0300)]
[IE TESTS][IE CPU] Add smoke prefix for broadcast SLT (#3158)

3 years ago[IE TESTS] Added test for NMS (#2830)
Maxim Andronov [Tue, 17 Nov 2020 06:33:19 +0000 (09:33 +0300)]
[IE TESTS] Added test for NMS (#2830)

3 years ago[CPU] added MergePermuteAndReorder optimization + added test (#2519)
Anton Voronov [Tue, 17 Nov 2020 06:04:49 +0000 (09:04 +0300)]
[CPU] added MergePermuteAndReorder optimization + added test (#2519)

3 years agoProtect reading external weights from up-dir (#3098)
Mateusz Bencer [Tue, 17 Nov 2020 04:16:39 +0000 (05:16 +0100)]
Protect reading external weights from up-dir (#3098)

* introduced path sanitizing

* added tests, folders structure clean-up

* fixed test

* fixed up-dir path handling

* improved sanitizing for windows

* fix Windows test

* move sanitize impl

3 years agoLSTM/GRU/RNN Sequences : support for seq_lengths input (#2788)
Ivan Tikhonov [Tue, 17 Nov 2020 04:04:20 +0000 (07:04 +0300)]
LSTM/GRU/RNN Sequences : support for seq_lengths input (#2788)

* sequences to ti transformations, support for seq_lengths input, update reference implemetations, add new tests

* fix python api, update sequences to ti transformation

* fix sequences to ti transformation

* Update sequences to TI transformation: fix reverse sequence support

* update single layer tests, fix TI reference impl, fix Sequences to TI transformations

* ngraph code style

* fix build

* fix ngraph python api

* resolver review comments, refactoring

* Resolve review remarks

* delete xfail

3 years agoSpecification to ONNX external data feature (#2939)
Mateusz Bencer [Tue, 17 Nov 2020 03:57:57 +0000 (04:57 +0100)]
Specification to ONNX external data feature (#2939)

* Added external weights files doc

* review remarks

* review remarks

* review remarks

Co-authored-by: Michał Karzyński <4430709+postrational@users.noreply.github.com>
* review remarks. part 2

Co-authored-by: Michał Karzyński <4430709+postrational@users.noreply.github.com>
* fixed info about nullptr

* Apply suggestions from code review

Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* review remarks. part 2

Co-authored-by: Michał Karzyński <4430709+postrational@users.noreply.github.com>
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
3 years agoUse std::transform instead of vector constructor (#3151)
Tomasz Socha [Tue, 17 Nov 2020 03:57:00 +0000 (04:57 +0100)]
Use std::transform instead of vector constructor (#3151)

3 years ago[IE TESTS] Remove extra dependency from Shared lib (#3006)
Irina Efode [Mon, 16 Nov 2020 19:29:02 +0000 (22:29 +0300)]
[IE TESTS] Remove extra dependency from Shared lib (#3006)

* [IE TESTS][IE CMAKE] Add cmake option for configuration to skip tests

* [IE TESTS] Remove extra dependency from IE tests shared lib

* Revert to add flag

3 years agoCorrected "-i" argument description (#3142)
Mikhail Ryzhov [Mon, 16 Nov 2020 18:48:06 +0000 (21:48 +0300)]
Corrected "-i" argument description (#3142)

* Corrected "-i" argument description

According to their usage in samples

3 years ago[IE CLDNN] Add asymmetric quantization support to fsv16 imad 1x1 convolution kernel...
Jedrzej Hajduczenia [Mon, 16 Nov 2020 17:56:19 +0000 (18:56 +0100)]
[IE CLDNN] Add asymmetric quantization support to fsv16 imad 1x1 convolution kernel (#2941)

3 years agoFix for Reduce extractors and normalizer (#3136)
Evgeny Lazarev [Mon, 16 Nov 2020 15:50:13 +0000 (18:50 +0300)]
Fix for Reduce extractors and normalizer (#3136)

* Fixed extractor for ONNX ReduceXXX operations and fixed ReduceAxisNormalizer transformation

* Unit test for ReduceAxisNormalizer transformation

3 years ago[IE][VPU]: ExpGenerateProposals: added temporary buffer allocation (#3093)
Alexey Ershov [Mon, 16 Nov 2020 15:03:19 +0000 (18:03 +0300)]
[IE][VPU]: ExpGenerateProposals: added temporary buffer allocation (#3093)

ExpGenerateProposals: temporary buffer allocation moved from firmware to blob
* Added tmpBuffer
* Changed the amount of allocated memory
* firmware update

Co-authored-by: Ivan Poliksenov <ivan.poliksenov@intel.com>
3 years ago[IE][VPU]: NMS - fix memory allocation bug (#3133)
Andrey Sokolov [Mon, 16 Nov 2020 14:52:27 +0000 (17:52 +0300)]
[IE][VPU]: NMS - fix memory allocation bug (#3133)

3 years agoUpdate tolerance values for ONNX model tests (#3131)
Mateusz Tabaka [Mon, 16 Nov 2020 14:23:20 +0000 (15:23 +0100)]
Update tolerance values for ONNX model tests (#3131)

3 years agosmall refactor samples setBatchSize (#3053)
Anton Romanov [Mon, 16 Nov 2020 14:20:54 +0000 (17:20 +0300)]
small refactor samples setBatchSize (#3053)

* small refactor samples setBatchSize

3 years agoExtend test coverage for ONNX Resize Operator (#3086)
Bartosz Sledz [Mon, 16 Nov 2020 14:20:01 +0000 (15:20 +0100)]
Extend test coverage for ONNX Resize Operator (#3086)

3 years agoAdd ceil_mode for Max and Avg pooling (#2965)
Bartosz Sledz [Mon, 16 Nov 2020 14:16:24 +0000 (15:16 +0100)]
Add ceil_mode for Max and Avg pooling (#2965)

3 years agoBump wheel from 0.34.2 to 0.35.1 in /ngraph/python (#3120)
dependabot[bot] [Mon, 16 Nov 2020 13:08:12 +0000 (14:08 +0100)]
Bump wheel from 0.34.2 to 0.35.1 in /ngraph/python (#3120)

3 years ago[CPU] Broadcast check has been added to the ConvertPowerToPowerIEMatcher transformati...
Maksim Kutakov [Mon, 16 Nov 2020 12:58:10 +0000 (15:58 +0300)]
[CPU] Broadcast check has been added to the ConvertPowerToPowerIEMatcher transformation. (#3132)

* [CPU] Added broadcast check to the ConvertPowerToPowerIE transformation.

* [CPU] Eltwise single layer test instance extended with a two inputs case.

3 years agoFixed negative paddings for convolution backprop data (#3117)
Ilya Churaev [Mon, 16 Nov 2020 12:30:18 +0000 (15:30 +0300)]
Fixed negative paddings for convolution backprop data (#3117)

* Fixed negative paddings for convolution backprop data

* Fixed code style

* Fixed comments