Commit Graph

422 Commits

Author SHA1 Message Date
Tomasz Jankowski
de4ceba375 [ONNX FE] Add implementation of Frontend API methods for naming and annotation (#8026) 2021-11-02 09:35:04 +01:00
Mang Guo
0a850ce73e [Shape infer] Implement ExperimentalDetectronTopKROIs shape infer (#7926)
* Implement ExperimentalDetectronROIs shape infer

* Fix review comments

* Update for review comments

Co-authored-by: Evgenya Stepyreva <evgenya.stepyreva@intel.com>
2021-11-02 05:32:46 +03:00
Ilya Churaev
4122ef50d6 Introduce OV Extension base api (#7562)
* Moved so loader to utils

* Fixed extension tests

* Fixed tests and style

* Fixed style and tests

* Fixed ARM build

* Fix windows

* Fix ieFuncTests

* Wrap runtime exception

* Fixed tests

* Added separate new extension

* Fixed unicode extension loading

* Try to fix windows

* Fixed windows

* Fixed macro

* Fixed doc

* Fixed build

* Fixed comments

* Try to fix build

* Fixed build

* Fixed build

* Fixed shared_from_this

* Temp commit

* Changed extension

* Fixed merge conflicts

* Removed ngraph namespace from new extensions

* Fixed code style

* Added core add_extension methods and tests

* Added new tests

* Implement tile operation

* Enabled new extensions support

* Fixed build

* Fixed code style

* Try to fix windows

* Changed base extension class

* Removed redundant Ptr

* Fixed comments

* Fixed friend decl

* Fixed Windows export

* Fixed centos

* Added template add_extension method

* Move destructor to public

* Removed BaseExtension class

* Added variadic add_extension methods

* Fixed doc and typo

* Added BaseOpDestructor

* Allow to create new extension only for new operations

* Revert tests

* Fixed comments

* Fixed comments

* Fixed comment

* Added SO Extension wrapper
2021-11-01 10:36:30 +03:00
Ilya Churaev
94206b9e2d Removed redundant evaluate methods (#8283)
* Removed redundant evaluate methods

* Added If to eval map

* Disable unsupported test cases
2021-11-01 09:08:32 +03:00
Evgenya Stepyreva
d50a40cd09 Shape inference (#8307)
* Convolution update

* Adds pragma once

* Reductions shape infer

* Shape nodes

* style
2021-11-01 05:39:57 +03:00
Piotr Szmelczynski
c729a6bfca [PYTHON] Expose Layout class (#7578)
* Create binding for Layout class

* add import of Layout to __init__

* add Layout to pyngraph.cpp

* create test for Layout

* fix typo

* fix style

* clean file

* fix style

* add Layot import to ngrapph init

* remove const char* init

* style fix

* move Layout from pyngraph/ to pyopenvino/graph

* style fix

* add doc() and args to Layout

* add Layout to openvino init

* fix style

* change import name

* fix import

* add __str__ impl

* remove duplicate

* style fix

* update test_layout

* add eq and ne operators

* create test for eq and ne operators

* fix style

* update copyright year
2021-10-28 19:46:50 +03:00
Ilya Lavrenov
8fb699a42c Static compilation for inference plugins (#8197)
* 1. Removed explicit SHARED from libraries
2. Fixed double definition for ie_layer_validators

* Fixed SEG in unit-test: order of initialization for global vars

* Added an ability to find plugins.xml from static IE

* Fixes in unit-test

* Migrated to new macro for import / export

* Minimized number of custom dllexport

* Don't use IR v7 for static libraries

* Revert for merge

* Don't enable tests with dlopen for static libraries

* Code style

* Added condition for export

* Revert format_reader

* Removed forward decalaration with external linkage

* Fixed IE linkage on Windows

* Reverted back 2 flags

* Minimal RRTI for cpuFuncTests

* Minimal RRTI for cpuFuncTests

* Still need IR v7 reader

* Fixed build

* Fixed compilation

* clang-format fix

* Removed BUILD_AS_IS and used USE_STATIC_IE

* Enable IR v7 reader as static library

* Fixed compilation for GPU plugin

* Trying to build plugins as static library

* Plugins are able provide their own name for CreatePluginEngine function

* Fixed CPU

* Fixed comments

* Fixed ENABLE_IR_V7_READER usage

* Fixed VPU

* clang-format

* Fixes

* Fix

* Load multiple plugins at once

* Fixed interpreter undefined symbols

* Trying to dynamically register static plugins

* Reverted some ngraph changes

* Fixed cpuUnitTests compilation

* Fixed compilation

* Fixed myriad

* Fixed custom_opset tests

* Reverted linker flags

* Support both static and dynamic plugins

* Fixed compilation of myriadFuncTests

* Removed duplication

* Fixes after self-review

* Fixed linkage for preprocessing

* Fixes for Windows

* Fixes

* Fixed cmake options

* Fix

* Fix

* Fix 2
2021-10-28 11:33:56 +03:00
Mikhail Nosov
5cb2174721 Layout: add _idx suffix for helper getters (#8196)
'layout::width' -> 'layout::width_idx', etc
2021-10-28 11:26:55 +03:00
Ilya Lavrenov
fe457aa59c Wmissing declarations for core OpenVINO (#8242)
* Fixed common libraries

* Enable flags only for ngraph, FEM, plugins and IE

* Fixed OpenVINOConfig.cmake

* Fixes for plugins

* Fixed link stage
2021-10-27 22:45:16 +03:00
Mikhail Nosov
846d1e0467 [OV2.0+Bugfix] Layout serialization + Fix caching issues with auto-generated friendly names (#8153)
* Fix caching issues with auto-generated friendly names

Introduce "HashPass" to calculate hash like a serialize but without auto-generated friendly names
IE Compilation context: use HashPass to calculate hash of ov::Function
IE Compilation context tests: removed "friendly names" from tests
Layout: serialization support + tests
Parameter/Result: update set/get_layout according to layout changes

* Fix clang

* Tiny correction of CmakeLists

* Renamed VariantWrapper<Layout> to LayoutAttribute
Removed test for conversion from/to fully dynamic layout (allowed now and does nothing)
'set_layout' - remove layout runtime info if empty layout is passed. This allows hashes to be the same if set_layout(param->get_layout()) is called

* Fix build

* change size_t to uint64_t for hash value
Using size_t leads to high probability of conflict for 32-bit platforms

* Removed debug print

* Fix hash calculation for 32-bit platforms

* Fix review comment
2021-10-27 21:05:06 +03:00
Ilya Lavrenov
e481758f73 Removed some global variables from ngraph (#8229) 2021-10-27 15:47:48 +03:00
Mikhail Nosov
d65e7d4d4f [OV20] Preprocessing: convert_layout<std::vector<uint64_t>> implementation (#8213)
* Preprocessing: convert_layout<std::vector<uint64_t>> implementation

User is able to use this version without specifying layout explicitly

Same version of convert_layout is added for post-processing

Added usage of new convert_layout to ie_network_reader

* Fix review comment
2021-10-27 11:32:44 +03:00
Anton Pankratv
c39e6fcfd8 added ov InferRequest behaviour tests (#7811) 2021-10-25 10:36:17 +03:00
Ilya Lavrenov
479081ff32 Minimized a number of static variables in static build (#8131)
* Minimal RRTI for cpuFuncTests

* Fixed build

* Fixed compilation

* clang-format fix

* Fixed compilation for GPU plugin

* Fixed CPU

* Fixed comments

* Fixes

* Fix

* Reverted some ngraph changes

* Fixed cpuUnitTests compilation
2021-10-25 10:23:04 +03:00
Ilya Churaev
838bd20065 Removed autogeneration of tensor names, added new reshape API (#8134)
* Removed autogeneration of tensor names, added new reshape API

* Fixed code style

* Fixed tests

* Fixed tests

* Fixed comments
2021-10-25 07:34:24 +03:00
Ilya Lavrenov
eb7f261d15 Several samples moved to OpenVINO 2.0 (#7970)
* ngraph functio sample

* Fixes

* OV2.0: NV12 hello classification with new preprocessing API

* Fix clang-format

* Rewrote several samples

* Some updates

* clang-format

* Fixed TODO

* Fixed comments and improved samples

* Preprocessing comments

* clang-format

* Fixed comment

* Fixed Windows

* Added back Blob support to ClassificationResult

Co-authored-by: Michael Nosov <mikhail.nosov@intel.com>
2021-10-23 22:47:35 +03:00
Ilya Lavrenov
dfefb92164 Made ov::runtime::Tensor strides to be in bytes (#8078) 2021-10-21 22:35:58 +03:00
Mikhail Nosov
d39fe50470 [OV20] Preprocessing: reverse_channels and RGB<->BGR conversion (#8098)
* Initial draft

* Support of dynamic shapes
2021-10-21 11:22:43 +03:00
Ilya Lavrenov
1b1e0e7553 preparation for static compilation (#8099)
* 1. Removed explicit SHARED from libraries
2. Fixed double definition for ie_layer_validators

* Fixed SEG in unit-test: order of initialization for global vars

* Added an ability to find plugins.xml from static IE

* Fixes in unit-test

* Migrated to new macro for import / export

* Minimized number of custom dllexport

* Don't use IR v7 for static libraries

* Revert for merge

* Don't enable tests with dlopen for static libraries

* Code style

* Added condition for export

* Revert format_reader

* Removed forward decalaration with external linkage

* Fixed IE linkage on Windows

* Reverted back 2 flags
2021-10-21 10:56:06 +03:00
Luwei Zhou
c2512f8dc5 [shape_infer]Implement ExperimentalDetectronDetectionOutput shape infer. (#7903)
* Implement ExperimentalDetectronDetectionOutput shape infer.

Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>

* Update on the review comments.

* Update based on review
2021-10-21 09:15:16 +03:00
Gleb Kazantaev
88f56669e2 Remove Provenance tags (#8101) 2021-10-20 10:28:58 +03:00
Ilya Churaev
4964ec890c Add add_output API for ov Function (#7980)
* Add add_output API for ov Function

* Fixed comments
2021-10-19 14:07:48 +03:00
Vladislav Golubev
c84db94697 [dynamism][CPU] MatMul transformations: dynamic shapes support (#7582)
* [TESTS] CPUUnitTests: unitTestUtils added to CMakeLists

* Reshape1D: functions moved to namespace

* [nGraph] PartialShape extending

* CPU specific transsformations for FullyConnected enabled to work with dynamic shapes

* [Transformations] FC transformations: dynamic shapes support

* [Transformations] FCBiasFusion: removed legacy check on bcast on weights

* [TESTS] FC transformations: tests

* [Transformations] SmartReshape: avoid tranformation utils methods

* codestyle fix

* codestyle fix

* [TESTS] MatMulTransformation tests: compilation error fix

* [CPU] FullyConnected: shape inference

* postreview fixes

* [CPU] FullyConnected: shape inference fix

* [nGraph] PShape dimensions insertion fixed

Co-authored-by: Stepyreva, Evgenya <evgenya.stepyreva@intel.com>
2021-10-19 10:59:29 +03:00
Ilya Lavrenov
b2977632d7 Fixed copy-paste in docs (#8046) 2021-10-15 19:53:16 +03:00
Katarzyna Mitrus
96df1a14ce Slice-8 Reference implementation (#7749)
* Init template reference tests

* Move output shape calculation to separate function

* Add to evaluate_map (reuse old ref)

* Add negative step test

* Make slice evaluate template type independent

* Enable typed tests for reference

* Add more reference tests

* Revert old slice ref changes

* New Slice reference implementation POC

* N-dim generalization

* Code refactor

* Style alignment

* Add more checks

* Add more tests

* Add zero dims/empty output tests

* Simplify typed tests generation

* Slice op leftovers

* Add unsigned types reference tests
2021-10-15 12:41:39 +03:00
Gleb Kazantaev
c1d2810321 Unify RT Info And Add Methods To Output class (#7871)
* Add methods to Output class

* Add get_any_name() method for tensor
2021-10-14 20:31:48 +03:00
Mikhail Nosov
d21572d7cb Squashed commit of the following: (#7930)
commit e692037384525e24f4f88d2b10fd5f5c09881d8a
Merge: a0ad24b16 82f8f19d1
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Wed Oct 13 15:06:20 2021 +0300

    Merge remote-tracking branch 'upstream/master' into ov20/remove_tensor_names

    # Conflicts:
    #	inference-engine/tests/functional/inference_engine/ir_serialization/rt_info_deserialization.cpp

commit a0ad24b16ab6ef7cf37dedb04084813178f0d627
Merge: 3c1f7067f db527fff4
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Wed Oct 13 13:45:39 2021 +0300

    Merge remote-tracking branch 'upstream/master' into ov20/remove_tensor_names

commit 3c1f7067f01634378cf66eee996d8793b635a6f2
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Wed Oct 13 13:45:13 2021 +0300

    Remove ignoring friendly names while comparing functions

commit f31e018d9a4694e83a3f12e685d8e90a9aca9045
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 23:50:20 2021 +0300

    Fix macos test

commit 2e3d0ceced3bde893e1e8237f7769c94bd193531
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 20:06:06 2021 +0300

    Fixes after merge

commit 5f047070af7c71a75f5ec019a146e17b3f95c062
Merge: 1568c9723 c323775f2
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 18:59:21 2021 +0300

    Merge remote-tracking branch 'upstream/master' into ov20/remove_tensor_names

    # Conflicts:
    #	inference-engine/tests/functional/inference_engine/ir_serialization/rt_info_deserialization.cpp
    #	inference-engine/tests/functional/plugin/cpu/shared_tests_instances/skip_tests_config.cpp
    #	ngraph/core/src/preprocess/pre_post_process.cpp
    #	ngraph/core/src/preprocess/preprocess_steps_impl.cpp
    #	ngraph/core/src/preprocess/preprocess_steps_impl.hpp
    #	ngraph/test/util/graph_comparator.cpp

commit 1568c97238c206a43cd8b14b6aba49fcee828386
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 13:56:48 2021 +0300

    Throw an error if 'InputInfo' is specified for same parameter twice.

commit d1830ba61b750b12484322c97261c7062551f798
Merge: 4f01b396b d79020457
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 13:34:32 2021 +0300

    Merge remote-tracking branch 'upstream/master' into ov20/remove_tensor_names

commit 4f01b396ba4b68d88d583092a38836754e78531c
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 13:34:19 2021 +0300

    Insert multi-plane parameters into new parameters vector (don't append them)

commit 70f85e5f351a37f2b79cff118399778c55136845
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Tue Oct 12 00:26:43 2021 +0300

    Fixed failed CPU/GPU test, created skipped test to clearly reproduce the problem
    Don't generate friendly and tensor names for sub-planes. Test is updated
    Keep the order of created parameters and results after pre and post processing
    Added tests for params/results order

commit 8b2cbf6db6646f6b2a01fc9e75726b682aa87fc2
Merge: 6050afbce 6d322722c
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Mon Oct 11 17:55:25 2021 +0300

    Merge remote-tracking branch 'upstream/master' into ov20/remove_tensor_names

commit 6050afbce256f21430322ec7b0a38105e3f06066
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Sun Oct 10 23:54:40 2021 +0300

    IE reader: remove unnecessary convert_layout()

commit 0f43133f6dfddf41233835678872c9a80b631565
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Sun Oct 10 23:33:29 2021 +0300

    Update IE tests

commit 75afe69ccf36b9764dc9285d77fe1a65ae4146b4
Merge: c86366577 67cfc9beb
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Fri Oct 8 22:52:17 2021 +0300

    Merge commit '67cfc9beb5bb0c66916c91ab5820dc25ad164a70' of https://github.com/openvinotoolkit/openvino into ov20/remove_tensor_names

commit c86366577ede8491ed722f9d048b7556d3266c68
Author: Michael Nosov <mikhail.nosov@intel.com>
Date:   Fri Oct 8 22:26:17 2021 +0300

    First commit (IE func tests are failed)

commit 67cfc9beb5bb0c66916c91ab5820dc25ad164a70
Author: y <ilya.lavrenov@intel.com>
Date:   Fri Oct 8 16:09:11 2021 +0300

    Removed explicit preprocessing steps
2021-10-14 17:54:41 +03:00
Ilya Churaev
eb838d5699 Cleanup Core API added common header (#7994) 2021-10-14 12:58:43 +03:00
Anton Pankratv
5b176f28cc Added remote objects type check (#7606) 2021-10-14 11:32:19 +03:00
Ilya Churaev
594da7e99d Fixed export of AttributeAdapter for broadcast spec (#7990) 2021-10-14 07:46:47 +03:00
Nikolay Shchegolev
0ed1c24cd2 [CPU] If-8 operation implementation. (#7253) 2021-10-13 20:45:24 +03:00
Ivan Tikhonov
7b1a418bf4 MakeStateful transformation (#7417)
* ReplaceInOutWithMemory transformation in ngraph

* add unit tests

* add ReplaceInputsOutputsWithMemory transformation in python

* update codestyle

* rename the transformation

* fix codestyle

* add Dynamic shapes check in the transformation and unit test

* fix codestyle

* rename files

* fix python API

* fix codestyle

* fix codestyle

* update python API

* fix codestyle

* fix build

* codestyle

* fix unit test

* fix RTTI declaration

* Apply suggestions from code review

Co-authored-by: Gleb Kazantaev <gleb.nnstu@gmail.com>

* review comments

* openvino codestyle

* change the name of Variable in the transformation

* fix build

* delete MakeStateful transformation from ie_transformations

* Resolve review comments

* codestyle

* fix missprint, codestyle

* delete unused variable

Co-authored-by: Gleb Kazantaev <gleb.nnstu@gmail.com>
2021-10-13 19:02:42 +03:00
Ilya Churaev
50b0dc1182 Deprecated old evaluate methods (#7937)
* Deprecated old evaluate methods

* Suppress node

* Fixed some warnings

* Fixed documentation and warnings

* Fixed some warnings

* Disable some warnings
2021-10-13 12:47:14 +03:00
Ilya Churaev
82f8f19d11 Moved serialization pass to public (#7862)
* Moved serialization pass to public

* Fixed unit tests

* Fixed code style

* Fixed tests

* Fix link

* Fix kmb

* Fixed build

* Removed InferenceEngine::Core from ngraph tests

* Add missed models

* Restore IR weights

* Fixed code style
2021-10-13 11:17:53 +03:00
Ilya Churaev
972524f1cc Use runtime::Tensor instead of Blob in read_model (#7951)
* Use runtime::Tensor instead of Blob in read_model

* Fixed the memory ownership
2021-10-13 09:13:11 +03:00
Mikhail Nosov
2592856200 [OV2.0] Leftovers for Preprocessing/input/output (#7893)
* Pre-process:
- Implicit conversions for element type and layout
- 'convert_element_type' with default argument to network
- Convert_element_type - don't add ops if dst and src types are same
- Convert_layout - don't add ops if dst and src layouts are same
- Custom step - use Output<Node> instead of shared_ptr<Node>
- Support of addressing input by tensor name

Post-process:
- Avoid duplication of tensor names after post-processing

* Fixed IE tests
2021-10-08 12:58:29 +03:00
Ilya Churaev
ff31ac553a Introduce new evaluate API (#7848)
* Introduce new evaluate API

* Removed redundant pointers from new API

* Fixed eval tests

* Fixed comments

* Enable evaluate_lover/upper methods
2021-10-08 07:29:25 +03:00
Mikhail Nosov
e20cefb620 PrePostProcessor.output() - first implementation of post-processing (#7866)
* PrePostProcessor.output() - first implementation of post-processing
Supported convert_layout, convert_element_type and custom operations

* Fix review comments

* Added test for pre and post processing together
Fix clang-format

* Move 'validate_and_infer_types' before post-processing
2021-10-06 21:17:10 +03:00
Mikhail Nosov
f57dc05c66 [OV20] Convert NV12 to RGB operation + preprocessing (#7508)
* # Conflicts:
#	docs/template_plugin/tests/functional/op_reference/convert_color_nv12.cpp
#	inference-engine/tests/functional/plugin/cpu/shared_tests_instances/single_layer_tests/convert_color_nv12.cpp
#	inference-engine/tests/functional/shared_test_classes/include/shared_test_classes/single_layer/convert_color_nv12.hpp
#	inference-engine/tests/functional/shared_test_classes/src/single_layer/convert_color_nv12.cpp
#	ngraph/core/include/openvino/core/preprocess/input_tensor_info.hpp
#	ngraph/core/include/openvino/core/preprocess/preprocess_steps.hpp
#	ngraph/core/include/openvino/op/nv12_to_bgr.hpp
#	ngraph/core/include/openvino/op/nv12_to_rgb.hpp
#	ngraph/core/src/op/nv12_to_bgr.cpp
#	ngraph/core/src/op/nv12_to_rgb.cpp
#	ngraph/core/src/preprocess/pre_post_process.cpp
#	ngraph/core/src/preprocess/preprocess_steps_impl.hpp
#	ngraph/test/CMakeLists.txt

* Added more test to cover 100% of code
Allow convert element type for 'multi-plane' color format

* Inherit tensor names for 'convert_color'

* Clang

* Fix tests

* Disable 'int8' preprocessing resize test

* Fix review comments

* Add more restrictions and tests for planes sub-names

* 1) Added check for uniqueness of tensor names generated for nodes
Raise error if user's plane sub-name conflicts with some node in a function
2) Added exception safety to preprocess build. Before, when input #2 fail, only one preprocess will be applied to function and it will be corrupted
Exception guard will restore function to original state if exception occurs

* Fix clang-format
2021-10-06 15:22:05 +03:00
Ilya Churaev
80e0bed855 Remove ngraph from new API (#7838) 2021-10-06 14:17:27 +03:00
Gleb Kazantaev
5ef3472bee Fix FrameworkNodeAttr Deserialization (#7854) 2021-10-06 12:40:26 +03:00
Ilya Churaev
3859990b1b Introduce IR v11 (#7787)
* Enabled support of IR v11 in IR frontend

* Enabled serialization to v11

* Added indexes for Inputs/Outputs

* Fixed typo

* Fixed code style

* Fixed deserialization tests

* FIxed typo

* Avoid changes in the original serialized function

* Fixed TI tests

* Removed version from the function and indexes from parameters/results

* Restore function version

* Clone function rt info

* Removed v11 version
2021-10-05 13:09:59 +03:00
Ilya Lavrenov
0bebd53b4a Inference chaining: static and dynamic cases (#7776)
* Fixed precisions conversion in new API

* Added tests

* Fixed old IR cases

* Disable FP16

* Fixed regex for CentoOS

* Refactored tests to use new API

* Temp

* Fixed tests

* Moved smart reshape related sources to ngraph

* Added tests for invalid names

* Moved reshape to tensor_names

* clang-format

* Fixed CC build

* Removed IEConv, IEDeconv from primitives pririty

* Added tests for Inference chaining

* Fixed dynamic chaining for template plugin

* Added test for 2 conflicting names for the single parameter

* Removed invalid test

* Added more tests for dynamism

* Fixed clang-format

* Fixed macosx compilation

* Some simplifications
2021-10-04 14:51:14 +03:00
Tomasz Jankowski
96519b1723 [nGraph] Remove obsolete attribute from Result op (#7682)
* Remove obsolete attribute from Result op

m_needs_default_layout field had no effect to the Result operator state,
even if set to true* by build_graph test.
Its setter and getter were unused.

* rather accidentaly considering implicit cast from const char*

* Restore redundant ctor as deprecated

It's user API so better keep it backward compatible

* Restore useless accessors as deprecated
2021-10-04 07:24:02 +03:00
Mikhail Nosov
5b39e407d9 [OV2.0] Preprocessing - add 'convert_layout' (#7772)
* Preprocessing - add 'convert_layout'

* Fix comments, removed 'set_data_shape' for time being
2021-10-01 19:45:40 +03:00
Ilya Churaev
faeaf045a9 Graph comparator to ngraph util (#7729)
* Moved FrameworkNode to nGraph

* Moved graph comparator to ngraph test util

* Fixed build

* Try to fix centos

* Fix export target

* Moved engine utils to separate library

* Removed ONNX util from common library

* Fixed build

* Fixed code style
2021-10-01 07:24:28 +03:00
Mikhail Nosov
6e05cead11 [OV20] Reference implementation for NV12toRGB and NV12toBGR operations (#7601)
* Reference implementation for NV12toRGB and NV12toBGR operations
Tests:
- ngraph: visitor + type_prop
- template plugin: reference implementation
- inference-engine: shared tests for plugins
- cpu plugin: compare with ref implementation tests

* Fix clang

* Serialization tests

* Fix clang-format

* Changed 'f32' to 'any supported floating-point type'
Added appropriate shape inference tests
Added error test for >2 inputs
Fixed failed CI tests

* Updates after rebase
+ Try to fix Ninja build

* Fix CI

* Support endianness + potential fix of win32 test fails

* Fix review comment

* Fix review comments

* Fix unit test build

* Fix unit test build #2

* Possible build fix 3

* Simplified reference tests
Observed issue with shuffling Y pixels on little-endian systems, added tests
2021-09-30 16:34:46 +03:00
Ilya Lavrenov
757db35528 Dynamism in new API (#7742)
* Fixed precisions conversion in new API

* Added tests

* Fixed old IR cases

* Disable FP16

* Fixed regex for CentoOS

* Refactored tests to use new API

* Temp

* Fixed tests

* Moved smart reshape related sources to ngraph

* Added tests for invalid names

* Moved reshape to tensor_names

* clang-format

* Fixed CC build

* Removed IEConv, IEDeconv from primitives pririty
2021-09-30 15:04:24 +03:00
Mikhail Nosov
414c3dc133 [OV2.0] Preprocessing - resize (#7583)
* Initial version

* Added 'network' layout to preprocessing info
Moved existing resize tests to template plugin

* Fix clang

* More tests for 'resize' reference implementation + CPU tests + error cases
Coverage is 100%

* Align with new base_reference_test implementation

* Fixed comments

* Add assert to check that desired size is not out of bounds

* CPU: skip failed test
2021-09-30 11:41:25 +03:00
Gleb Kazantaev
62ed238f54 RT Info Attributes Serialization/Deserialization (#7603)
* Init rt info deserialization

* Add RT Info attributes deserialization

* Add RT Info attributes serialization

* Add Serialization and Deserialization tests

* Code style

* Fix AttributeAdapter place

* Extended rt_info section; added support for multi-field attrs; use version for attr serialization/deserialization

* Align attributes and their usage

* Fix LPT

* Add missing #pragma one

* Fix build

* Change version delimiter to have valid xml

* Fix PromitivesPriority

* Fix tests

* Fix test

* Fix IR Reader; remove empty rt_info

* More explicit attribute name and version representation
2021-09-29 19:17:03 +03:00