IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
5,441 Commits 38 Branches 54 Tags
2f612f776be977b69897176f9bfc54aca20a8b13
Commit Graph

3238 Commits

Author SHA1 Message Date
Maxim Vafin
2f612f776b Define common frontend (#8110) 
* Rename and reorganize frontend_manager

* Apply codestyle

* Fix installing headers

* Fix codestyle
 2021-11-24 18:56:05 +03:00
Irina Efode
27fa64031b [IE TESTS] Add QeryNet to conformance (#8768) 2021-11-24 15:54:03 +03:00
Edward Shogulin
0ba105e8af [CPU] DetectionOutput dynamism enabling (#8576) 2021-11-24 15:10:17 +03:00
Anton Dudchenko
2da56312e6 [VPU] Enable some tests (#8438) 
Removed several tests from skip list, as they currently work on U20
 2021-11-24 14:17:04 +03:00
Mateusz Tabaka
2a9d3ab95f [GPU] recalculate output layout after pre_optimize_bias (#8742) 2021-11-24 13:43:23 +03:00
Polina Brzezinskaya
acaddf73a6 [VPU] Add support for dynamic shape API (#8470) 
This PR adds support for new dynamic shape API to MyriadX
 2021-11-24 13:22:05 +03:00
Anton Dudchenko
d33033bb67 [VPU] Not unload MyriadPlugin at runtime on ARM (#8772) 
* [VPU] Not unload MyriadPlugin at runtime on ARM

* Remove ARM definition
 2021-11-24 13:14:40 +03:00
Luwei Zhou
71b3007a79 [CPU] Enable space_to_depth dynamic shape support. (#8639) 2021-11-24 12:11:15 +03:00
Vladimir Gavrilov
edb98aeb8e nGraph version of the MO transformation InterpolateSequenceToInterpolate (#8397) 
* Started to write transformation to fuse Interpolate sequence.

* Some changes.

* Written the transformation to fuse two Interpolate layers.

* Deleted commented code.

* Small fixes.

* Some fixes.

* Started to write tests.

* Small fix.

* Added more tests. Deleted commented code.

* Deleted redundant headers.

* Small fix.

* Fixes in the function can_be_fused(): the last statement was decomposed.

* Added operators == and != for op::v4::Interpolate::InterpolateAttrs.

* Added more checks for nullptr.

* Fixed codestyle.

* Added Interpolate registration.

* Small change.

* Implementation of operator== of InterpolateAttrs was moved into header file.
 2021-11-24 11:33:09 +03:00
Alexandra Sidorova
24aed538e3 [CPU] Added dynamism support for NormalizeL2 (#8580) 2021-11-24 10:28:42 +03:00
Min, Byungil
2ecb2a1895 Remove unnecessary reorder to improve perf (#8341) 
+ Bugfix for invalid removal of reorder in front of fsv32 conv
+ Remove unnecessary reorder of first conv of bfyx output format

Signed-off-by: Min, Byungil <byungil.min@intel.com>
 2021-11-24 16:12:14 +09:00
Mateusz Tabaka
585c6bcb24 [CPU] Don't fuse Add or Mul to FakeQuantize if the result is not supported b… (#8585) 2021-11-24 09:50:47 +03:00
Taylor Yeonbok Lee
abbf0384ae [GPU] Add GetMaxBatchSize functionality (#8226) 2021-11-24 09:30:14 +03:00
Sungeun Kim
8f64b8e2e8 [GPU] Update flops calculation for DG2/ATS (#8758) 2021-11-24 09:28:22 +03:00
Maksim Shabunin
94b6f64e0b oneDNN: fixed build with mullti-config generators (#8775) 2021-11-24 09:11:24 +03:00
Taylor Yeonbok Lee
32052a8f92 [GPU] Show freed constants' name on Verbose memory dump (#8354) 2021-11-24 10:58:58 +09:00
Vladimir Paramuzov
6addc0d535 [GPU] USM sharing and host blob creation in gpu remote context (#8657) 2021-11-23 20:26:44 +03:00
Alexandra Sidorova
c49620bb6a [CPU] Added dynamism support for ShuffleChannels (#8601) 2021-11-23 17:22:33 +03:00
Nikita Demashov
6156626706 [LPT] Fix Yolo v5 bug with Concat transformation subtract constant (#8491) 
* add supporting convert after subtract constant in concat_transformation

* add check for the existence of subtract constant

* add convert fold for subtract convert & add tests for subtract with convert
 2021-11-23 17:10:14 +03:00
Luo Cheng
fa65018773 [CPU] GatherElements support dynamic shape (#8663) 2021-11-23 16:53:04 +03:00
Luwei Zhou
64459177e7 [CPU] Implement spaceToBatch node dynamic shape support (#8701) 2021-11-23 16:10:26 +03:00
Egor Shulman
9ce8ac536f [CPU] Optimized shapeInfer() for OneHot (#8739) 2021-11-23 15:55:52 +03:00
Mikhail Nosov
c29569ecbd [TemplateFuncTests] RGBX->RGB: Increase image size to avoid test failures on some platforms (#8762) 
* Increase image size to avoid test failures on some platforms

There is an assert 'length >= nlanes' and 'nlanes' value depends on machine architecture
Set 320x320 test image to guarantee that image size >= nlanes

Also increased image size for 'plugin shared tests' to ensure the same (even though tests do not work with legacy preprocessing)

* Descreased to 160x160
 2021-11-23 15:26:22 +03:00
Alexandra Sidorova
8c55c761c4 [CPU] Fixed leftovers for GatherND (#8644) 2021-11-23 15:11:58 +03:00
Daria Mityagina
277ff96564 [IE][VPU][XLink]: Port changes from releases/2021/4 to master (#8404) 2021-11-23 12:51:46 +03:00
Elizaveta Lobanova
6feb981578 [GNA] Fixed handling of transposes around MatMul (#8600) 
* [GNA] Fixed handling of transposes around MatMul

* [GNA] Fixed swap matmul inputs tests

* [GNA] Comments applying
 2021-11-23 12:43:04 +03:00
Vladimir Paramuzov
24675fd8ea [GPU] Fixed a couple of issues in clDNN found by static analysis (#8706) 2021-11-23 10:33:03 +03:00
Mikhail Nosov
7545af07d0 Preprocessing: convert color RGBX/BGRX -> RGB/BGR implementation (#8755) 2021-11-23 10:29:00 +03:00
Zlobin Vladimir
251883001c [CPU] Set lower bound for nstreams to 1 (#8747) 2021-11-23 06:42:12 +03:00
Min, Byungil
c2ddfdc940 [GPU] Bugfix for accruacy regression of onednn reorder fusing (#8441) 
Signed-off-by: Min, Byungil <byungil.min@intel.com>
 2021-11-23 11:26:46 +09:00
Alexandra Sidorova
75af8254cb [CPU] Fixed performance for Eltwise with C=1 (#8406) 2021-11-23 01:09:52 +03:00
Mikhail Letavin
a862331732 [GPU] Fix multi configuration build of onednn_gpu (#8743) 2021-11-23 00:22:05 +03:00
Mikhail Nosov
d15dcd6833 Support 'set_memory_type(string)' for preprocessing info tensor info (#8721) 
Testing: unit tests + serialization/deserialization tests
 2021-11-22 19:42:23 +03:00
Gleb Kazantaev
9f54641858 Update Transpose Sinking (#8720) 
* Update Transpose Sinking

* Fix namespace

* Fix typo
 2021-11-22 16:47:00 +03:00
Vladimir Paramuzov
b05d90032b [GPU] Replace input_offset parameters with padding (#7571) 2021-11-22 14:45:58 +03:00
Szymon Irzabek
f6a4dcb5ac [GNA] Add MVN decomposition (#8142) 2021-11-22 14:39:25 +03:00
Sergey Shlyapnikov
a4854fa61d [GPU] Add memory benchmarking tests (#8689) 2021-11-22 14:25:05 +03:00
FanJiangIntel
de639899a6 Fix THREADING=SEQ build failure on Windows (#8699) 2021-11-22 12:54:32 +03:00
Vladislav Golubev
a33b75871d [CPU] Security fixes (#8456) 2021-11-22 10:08:37 +03:00
Yuan Hu
64e45447f4 fix compile issue caused by interface change (#8727) 
Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>
 2021-11-22 10:07:18 +03:00
Mikhail Ryzhov
390782764c [GNA] Enabled new OV inference tests (#8618) 
* Enabled the new OV inference tests

* reverted .gitignore

* Revert "reverted .gitignore"

This reverts commit a6b63bfef8.

* Update inference-engine/tests/functional/plugin/gna/shared_tests_instances/skip_tests_config.cpp

Co-authored-by: Nadezhda Ageeva <nkogteva@gmail.com>

* Update inference-engine/tests/functional/plugin/gna/shared_tests_instances/skip_tests_config.cpp

Co-authored-by: Nadezhda Ageeva <nkogteva@gmail.com>

* Removed CVS from skip list

Co-authored-by: Nadezhda Ageeva <nkogteva@gmail.com>
 2021-11-21 16:15:53 +03:00
Maxim Andronov
7c525592e7 [CPU] clone Constant node instead direct usage (#8719) 2021-11-21 14:50:17 +03:00
Egor Shulman
c4e49bb915 [CPU] OneHot dynamic done (#8488) 2021-11-21 14:44:11 +03:00
Ilya Lavrenov
a1e95f4d69 Deprecated SOPointer (#8711) 
* Removed SOPointer

* Updates

* Fixed tests, compilation
 2021-11-20 18:13:10 +03:00
Yuan Hu
063df47ec9 continue to load candidate device when load device failed (#8006) 
* fix potential risk

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* try catch exception when load nextwork

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* wait for the first ready device instead of cpu

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* continue to select next device if load failed

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* Refactoring code to make code easy and readable

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* debug core dump

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* improve the code

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* add test code

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* test draft

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* improve test code

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* remove debug print

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix some reviews

add comments on test
init flags in hpp
remove draft code

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix code format

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* try to fix compile static library issue in ci

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix compile issue in centos76

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix build issue on windows

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix a review

merge loop

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* rebase master

fix conflict with PR 8389

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix some review issue

lock issue
change NULL to nullptr
misprint issue
add EXPECT_CALL for ON_CALL

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* run test in ci

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* add more EXPECT_CALL test

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* fix the comment of test config

Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>

* Fix merge from master (was required to resolve conflicts)

Co-authored-by: Maxim Shevtsov <maxim.y.shevtsov@intel.com>
 2021-11-20 17:49:19 +03:00
Ilya Lavrenov
543ca4f318 Enabled proper OpenVINOConfig.cmake generation for static build (#8634) 
* Enabled PDPD, TF on Windows

* Try to build samples against static libs

* Fixes

* Fixes

* fix

* Used ENABLE_GAPI_PREPROCESSING in C API tests

* Revert "Used ENABLE_GAPI_PREPROCESSING in C API tests"

This reverts commit df9d53e3cc.

* Fixes

* Revert exclude

* Enabled proper OpenVINOConfig.cmake generation

* Try

* Try
 2021-11-20 02:27:43 +03:00
Anton Pankratov
bcb7415fce Fixed exec net config test (#8710) 2021-11-19 23:31:42 +03:00
Gleb Kazantaev
0e749b8b15 Reduce number of Function Validations (#8668) 2021-11-19 13:04:49 +03:00
song, bell
ea0095ad6d stick to default memory if userrequest exceed the optimal number (#8640) 
* stick to default memory if userrequest exceed the optimal number

Signed-off-by: fishbell <bell.song@intel.com>

* formatting

Signed-off-by: fishbell <bell.song@intel.com>
 2021-11-19 10:29:16 +03:00
Mikhail Nosov
aaac5de2fa [OV2.0] I420 preprocessing color format support (#8680) 
* Add I420 preprocessing

Tests:
- Template reference
- OpenCV
- Legacy Preprocessing tests
- Functional shared tests for plugins (CPU)
- Disable i420 preprocessing tests on GPU

* clang-format fix
 2021-11-18 23:30:44 +03:00