platform/core/ml/nnfw.git
4 years ago[nnfw-cmake] Revise build option setting (#7780)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 27 Sep 2019 02:13:08 +0000 (11:13 +0900)]
[nnfw-cmake] Revise build option setting (#7780)

- Reorder correct place for default in CfgOptionFlags.cmake
- Remove redundant platform specific setting - same with default
- Remove redundant option setting in tizen cross build script
- Remove BUILD_TOOLS option
- Introduce BUILD_OPENCL_TOOL to set build option for each tool

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[nnfw-cmake] Replace boost package config (#7786)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 27 Sep 2019 00:18:14 +0000 (09:18 +0900)]
[nnfw-cmake] Replace boost package config (#7786)

Move boost & source package config to common directory

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[nnfw-cmake] Replace acl package config (#7785)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 27 Sep 2019 00:17:30 +0000 (09:17 +0900)]
[nnfw-cmake] Replace acl package config (#7785)

Move acl & source package config to common directory

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[exo-tflite] Introducing TFLTranspose (#7783)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Thu, 26 Sep 2019 23:39:41 +0000 (08:39 +0900)]
[exo-tflite] Introducing TFLTranspose (#7783)

TFLTranspose is added to represent Tensorflow Lite Transpose operation.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[loco] Introducing FilterDecode (#7778)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Thu, 26 Sep 2019 11:48:06 +0000 (20:48 +0900)]
[loco] Introducing FilterDecode (#7778)

This adds FilterDecode, which converts a filter to a tensor.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[exo-tflite] Refactoring ReluConverterTest (#7781)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Thu, 26 Sep 2019 11:46:49 +0000 (20:46 +0900)]
[exo-tflite] Refactoring ReluConverterTest (#7781)

1) using TestGraph to shorten code; 2) first test does not check loco::Pull

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[tf2tflite] Enable NET_0028 test (#7767)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 11:33:48 +0000 (20:33 +0900)]
[tf2tflite] Enable NET_0028 test (#7767)

Let's add NET_0028 to tf2tflite integration testsuite.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[moco] Enable NET_0028 test (#7766)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 11:33:29 +0000 (20:33 +0900)]
[moco] Enable NET_0028 test (#7766)

Let's add NET_0028 to moco integration testsuite.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[Dist/Tizen] Do not install libraries with devel packages. (#7769)
함명주/On-Device Lab(SR)/Principal Engineer/삼성전자 [Thu, 26 Sep 2019 09:50:35 +0000 (18:50 +0900)]
[Dist/Tizen] Do not install libraries with devel packages. (#7769)

Do not install duplicated libraries with devel packages; it incur build errors in applications with indirect dependencies on nnfw.
For example, if A Requires B and B BuildRequires nnfw, building A indirect-BuildRequires nnfw's *.so files, which cannot be resolved by GBS or OBS because both nnfw and nnfw-devel have nnfw's *.so files.
Thus, if nnfw has all *.so files in _libdir, do not install them with devel except for representative symlinks.

Fixes #7714

Signed-off-by: MyungJoo Ham <myungjoo.ham@samsung.com>
4 years agoRevise tizen build (#7755)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 09:46:37 +0000 (18:46 +0900)]
Revise tizen build (#7755)

Remove special setting for gbs/obs build
Install armcl package in tizen rootfs
Move default setting for tizen into cmake option

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[tfldump] Use external TensorFlow 1.14 schema file (#7762)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Thu, 26 Sep 2019 08:29:19 +0000 (17:29 +0900)]
[tfldump] Use external TensorFlow 1.14 schema file (#7762)

This commit uses tflite schema file from TensorFlow 1.14 external source
code.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[moco-tf] Remove Phase (#7760)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 07:28:12 +0000 (16:28 +0900)]
[moco-tf] Remove Phase (#7760)

logo replaces this internal implementation of Phase.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[moco-tf] Introduce CodecHelper (#7757)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 07:27:44 +0000 (16:27 +0900)]
[moco-tf] Introduce CodecHelper (#7757)

This commit introduces CodecHelper module which provides
set_feature_enc/set_feature_dec helpers.

This change reduces duplication in code base.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[loco] Remove unnecessary friend declarations (#7753)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 07:27:27 +0000 (16:27 +0900)]
[loco] Remove unnecessary friend declarations (#7753)

This commit removes unnecessary friend declarations from loco Graph
class.

These friend declarations are introduced for migration, but no longer
required.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[CI] Exit with fail code when build fail (#7758)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 07:23:57 +0000 (16:23 +0900)]
[CI] Exit with fail code when build fail (#7758)

Fix build script to check build fail

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[nncc-cmake] Replace package config files (#7761)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 06:54:58 +0000 (15:54 +0900)]
[nncc-cmake] Replace package config files (#7761)

Replace package config files and related files to common directory with nnfw

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[moco-tf] Reduce Code Duplication (#7743)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 05:50:34 +0000 (14:50 +0900)]
[moco-tf] Reduce Code Duplication (#7743)

This commit rewrites canonizliazation passes to use SimpleNodeTransform
in order to reduce code duplication.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[nncc-cmake] Rename nncc_find_package to nnas_find_package (#7740)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 04:20:15 +0000 (13:20 +0900)]
[nncc-cmake] Rename nncc_find_package to nnas_find_package (#7740)

Rename nncc_find_package to nnas_find_package to share package config with nnfw

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[tflite_bencmark] Let users can turn On/Off build (#7750)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Thu, 26 Sep 2019 02:44:47 +0000 (11:44 +0900)]
[tflite_bencmark] Let users can turn On/Off build (#7750)

* [tflite_bencmark] Let users can turn On/Off build

Introduce CMake variable `BUILD_TFLITE_BENCHMARK` which is a flag to
indicate do the build or not.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
* Update CfgOptionFlags.cmake

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[ARMComputeEx] Fix target_link_libraries dependency (#7745)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Thu, 26 Sep 2019 01:36:23 +0000 (10:36 +0900)]
[ARMComputeEx] Fix target_link_libraries dependency (#7745)

This library actually needs `arm_compute` rather than
`arm_compute_core`. Without this, Android NDK build fails.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[nnapi-test] Update test generator (#7729)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Thu, 26 Sep 2019 01:14:39 +0000 (10:14 +0900)]
[nnapi-test] Update test generator (#7729)

- Rename directory: andrdoid-10
- Use generator in android-10 r2 release (not used yet)
- Remove useless files
- Update README.md

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[android_tflite] Disable JNI build (#7744)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Thu, 26 Sep 2019 01:08:59 +0000 (10:08 +0900)]
[android_tflite] Disable JNI build (#7744)

JNI build is not trivial due to the reason that is described in the
comment. So temporarily disable JNI build.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[tflite_benchmark] Remove Boost dependency (#7751)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Thu, 26 Sep 2019 01:03:47 +0000 (10:03 +0900)]
[tflite_benchmark] Remove Boost dependency (#7751)

Accumulator is the only module that `tflite_benchmark` is depend on.
This commit introduces our own Accumulator and replace with it.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[nnfw-cmake] Introduce nnas_find_package (#7741)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Wed, 25 Sep 2019 23:27:00 +0000 (08:27 +0900)]
[nnfw-cmake] Introduce nnas_find_package (#7741)

Introduce nnas_Find_package to share package config with nncc

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoMake layout of frontend and backend to be the same for Gather, Reshape and FC (#7711)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 21:51:40 +0000 (06:51 +0900)]
Make layout of frontend and backend to be the same for Gather, Reshape and FC (#7711)

This commit makes layout of frontend and backend to be the same for Gather, Reshape and FC.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[mir] Add function for numpy-style broadcasting (#7749)
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 25 Sep 2019 19:00:23 +0000 (22:00 +0300)]
[mir] Add function for numpy-style broadcasting (#7749)

* Add `broadcastShapes` for numpy-style bidirectional broadcasting.
* Use it in shape inference of binary operations.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
4 years ago[nnc] Update C++-generating backend elementwise operations (#7748)
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 25 Sep 2019 15:43:52 +0000 (18:43 +0300)]
[nnc] Update C++-generating backend elementwise operations (#7748)

Remove support for multiple inputs to elementwise operations (they now accept only two inputs).

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
4 years ago[mir2loco] Implemented transformation for FullyConnectedOp (#7414)
Павел Ильютченко/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 25 Sep 2019 14:50:29 +0000 (17:50 +0300)]
[mir2loco] Implemented transformation for FullyConnectedOp (#7414)

* Support transformation mir::FullyConnected to loco::MatrixMul

Signed-off-by: Pavel Iliutchenko <p.iliutchenk@samsung.com>
4 years ago[nnfw-cmake] Use android cmake variable (#7738)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Wed, 25 Sep 2019 10:28:20 +0000 (19:28 +0900)]
[nnfw-cmake] Use android cmake variable (#7738)

Use android cmake variable ANDROID defined in toolchain cmake setting

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[nnfw-cmake] Remove nnfw module directory (#7720)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Wed, 25 Sep 2019 10:01:59 +0000 (19:01 +0900)]
[nnfw-cmake] Remove nnfw module directory (#7720)

Move remain nnfw cmake module to unified module
Remove nnfw cmake module
Use unified module in nnfw cmake script

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[android_tflite] Move CMake dependency declaration (#7725)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 08:22:41 +0000 (17:22 +0900)]
[android_tflite] Move CMake dependency declaration (#7725)

Move CMake dependency declaration to the correct file. Adding dependency
from another module is illegal.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[loco] Fix typo in CanonicalShapeInferenceRule (#7734)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Wed, 25 Sep 2019 07:56:06 +0000 (16:56 +0900)]
[loco] Fix typo in CanonicalShapeInferenceRule (#7734)

This will fix typo in CanonicalShapeInferenceRule

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[exo-tfltie] turning on a knob of TFLConst (#7733)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Wed, 25 Sep 2019 07:40:30 +0000 (16:40 +0900)]
[exo-tfltie] turning on a knob of TFLConst (#7733)

This turns on a knob for TFLConst.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[exo-tflite] Give name to created tensor (#7732)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 07:38:23 +0000 (16:38 +0900)]
[exo-tflite] Give name to created tensor (#7732)

* [exo-tflite] Give name to created tensor

This commit amends operator export stage to give name to just created
tensor.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
* Remove duplicated definition

4 years ago[moco-tf] SimpleNodeTransform (#7723)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Wed, 25 Sep 2019 04:54:47 +0000 (13:54 +0900)]
[moco-tf] SimpleNodeTransform (#7723)

This commit introduces SimpleNodeTransform which iterates over active
nodes of a certain type.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[tf2tflite] NET_0016 test (#7727)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 04:54:33 +0000 (13:54 +0900)]
[tf2tflite] NET_0016 test (#7727)

This commit adds NET_0016 test for tf2tflite

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[moco-integration-test] Use i5diff (#7728)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Wed, 25 Sep 2019 04:54:16 +0000 (13:54 +0900)]
[moco-integration-test] Use i5diff (#7728)

Let's use i5diff instead of h5diff.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[neurun] Add missing inclusions of headers (#7724)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 04:06:35 +0000 (13:06 +0900)]
[neurun] Add missing inclusions of headers (#7724)

Add inclusions for Android NDK build. NDK clang compiler complains
without this.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[tf2tflite] Conv2DBackpropInput unit tests (#7718)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 01:37:23 +0000 (10:37 +0900)]
[tf2tflite] Conv2DBackpropInput unit tests (#7718)

This commit adds following unit tests for tf2tflite integration test:
- UNIT_Conv2DBackpropInput_000, valid padding case
- UNIT_Conv2DBackpropInput_001, same padding case

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years agoMake backends of acl support PermuteNode (#7685)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 01:36:38 +0000 (10:36 +0900)]
Make backends of acl support PermuteNode (#7685)

This commit makes backends of acl support PermuteNode.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years agoMake to support SpaceToDepth op for acl neon (#7633)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 01:36:00 +0000 (10:36 +0900)]
Make to support SpaceToDepth op for acl neon (#7633)

This commit makes to support SpaceToDepth op for acl neon.
  - Introduce NESpaceToDepthLayerEx and NESpaceToDepthLayerKernelEx
  - Apply NEDepthToSpaceLayerEx layer for neurun

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[libs/tflite] Fix narrowing conversion (#7717)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Wed, 25 Sep 2019 01:35:29 +0000 (10:35 +0900)]
[libs/tflite] Fix narrowing conversion (#7717)

There is a narrowing conversion in `RandomGenerator` constructor and it
is detected by Android NDK clang compiler.

`int` to `uint_fast32_t` causes narrowing conversion so this commit
changes type of `seed` to `uint32_t`.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years agoFix nnfw.h application build errors (#7716)
함명주/On-Device Lab(SR)/Principal Engineer/삼성전자 [Tue, 24 Sep 2019 23:46:34 +0000 (08:46 +0900)]
Fix nnfw.h application build errors (#7716)

Fix build errors of applications using nnfw.h:
```
[   15s] /usr/include/nnfw.h:162:35: error: unknown type name 'nnfw_tensorinfo'
[   15s]                                    nnfw_tensorinfo *tensor_info);
[   15s]                                    ^~~~~~~~~~~~~~~
[   15s] /usr/include/nnfw.h:174:36: error: unknown type name 'nnfw_tensorinfo'
[   15s]                                     nnfw_tensorinfo *tensor_info);
[   15s]                                     ^~~~~~~~~~~~~~~
```

Fixes #7715

Signed-off-by: MyungJoo Ham <myungjoo.ham@samsung.com>
4 years ago[nnfw-cmake] Use unified cmake module in toolchain setting (#7683)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 06:44:41 +0000 (15:44 +0900)]
[nnfw-cmake] Use unified cmake module in toolchain setting (#7683)

Use unified function envoption in toolchain setting

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoMake to support HashTableLookup op for acl neon (#7694)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 05:21:01 +0000 (14:21 +0900)]
Make to support HashTableLookup op for acl neon (#7694)

This commit makes to support HashTableLookup op for acl neon.
  - Introduce NEHashTableLookup and NEHashTableLookupKernel
  - Apply NEHashTableLookup layer for neurun

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years agoUpdate arm64-android toolchain file (#7684)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 05:20:37 +0000 (14:20 +0900)]
Update arm64-android toolchain file (#7684)

Make the toolchain file to use the toolchain file that NDK privides.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[nnfw-cmake] Rename nnfw_include (#7681)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 04:15:58 +0000 (13:15 +0900)]
[nnfw-cmake] Rename nnfw_include (#7681)

Rename nnfw_include to nnas_include: prepare to use unified cmake module

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoAdd tf2tflite-model-test project (#7698)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 02:20:24 +0000 (11:20 +0900)]
Add tf2tflite-model-test project (#7698)

* Add tf2tflite-model-test project

Let's allows users to verify tf2tflite with their own models.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
* Add .gitignore

4 years agoIntroduce build_android_ndk.sh (#7675)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 02:10:13 +0000 (11:10 +0900)]
Introduce build_android_ndk.sh (#7675)

This commit introduces `build_android_ndk.sh` which sets up Android
build environment. And it will replace `build_android_rootfs.sh`.

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years agoAdd REGRESSION_0001 TensorFlow test (#7695)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 01:57:54 +0000 (10:57 +0900)]
Add REGRESSION_0001 TensorFlow test (#7695)

This commit introduces REGRESSION_0001 T/F test which includes two
successive broadcasted add.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[tf2tflite] Enable UNIT_Add_004 test (#7697)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 01:57:39 +0000 (10:57 +0900)]
[tf2tflite] Enable UNIT_Add_004 test (#7697)

Now, tf2tflite unittest suite includes UNIT_Add_004 by default.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[moco-integration] Conv2DBackpropInput unit tests (#7702)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 01:57:29 +0000 (10:57 +0900)]
[moco-integration] Conv2DBackpropInput unit tests (#7702)

This commit adds following unit tests for moco integration:
- UNIT_Conv2DBackpropInput_000, valid padding case
- UNIT_Conv2DBackpropInput_001, same padding case

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[moco] Add UNIT_Add_004 to integration test (#7709)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Tue, 24 Sep 2019 01:57:15 +0000 (10:57 +0900)]
[moco] Add UNIT_Add_004 to integration test (#7709)

Let's add UNIT_Add_004 as a part of moco integration test.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[exo-tflite] Export TransposedConv2D (#7707)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 01:57:06 +0000 (10:57 +0900)]
[exo-tflite] Export TransposedConv2D (#7707)

This commit supports canonical TransposedConv2D node to be exported as
TFlite TRANSPOSE_CONV operation.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[moco-tf] Update comment on eltwise binary canonicalization (#7704)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Tue, 24 Sep 2019 01:28:39 +0000 (10:28 +0900)]
[moco-tf] Update comment on eltwise binary canonicalization (#7704)

* [moco-tf] Update comment on eltwise binary canonicalization

This commit updates outdated comment on canonicalize_eltwise_binary_node().
It now can dealt with broadcast case.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
* Clarify comment

4 years ago[moco-tf] Check node shape for BinaryCanonicalze (#7703)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Tue, 24 Sep 2019 01:21:44 +0000 (10:21 +0900)]
[moco-tf] Check node shape for BinaryCanonicalze (#7703)

This will add check for node shape in EltwiseBinaryCanonicalzeHelper as this requires nodes shape

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[moco-tf] Remove SquaredDifferenceCanonicalizer (#7696)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 23:33:35 +0000 (08:33 +0900)]
[moco-tf] Remove SquaredDifferenceCanonicalizer (#7696)

This will remove SquaredDifferenceCanonicalizer that is not used.
Add a note for SquaredDifference node canonicalization.

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[mir2loco] DeConv2D transformation implementation (#7642)
Павел Ильютченко/AI Tools Lab /SRR/Engineer/삼성전자 [Mon, 23 Sep 2019 15:30:12 +0000 (18:30 +0300)]
[mir2loco] DeConv2D transformation implementation (#7642)

* Supported mir::DeConv2D transformation to loco::TransposedConv2D
* Added test for it

Signed-off-by: Pavel Iliutchenko <p.iliutchenk@samsung.com>
4 years ago[mir2loco] Remove no longer unused functions (#7666)
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Mon, 23 Sep 2019 12:01:19 +0000 (15:01 +0300)]
[mir2loco] Remove no longer unused functions (#7666)

Remove shape based `setupWindow` and `setupStride` which are no longer used.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
4 years ago[neurun]HEScheduler ignore backend if not profiled (#7641)
Дилшоджон Пошшоев/AI Tools Lab /SRR/Engineer/삼성전자 [Mon, 23 Sep 2019 09:59:13 +0000 (12:59 +0300)]
[neurun]HEScheduler ignore backend if not profiled (#7641)

Now HEScheduler throws an exception if there is no profiling info for an
operation on a backend. This change makes HEScheduler not use this
backend for that operation

Signed-off-by: Poshshoev Dilshodzhon <d.poshshoev@samsung.com>
4 years ago[nike] Use GTest_AddTest helper (#7689)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 09:15:57 +0000 (18:15 +0900)]
[nike] Use GTest_AddTest helper (#7689)

This commit revises nike's CMakeLists.txt to use GTest_AddTest helper
instead of explicitly referring "gtest_main" target.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[moco-tf] Same padding case for Conv2DBackpropInput Canonicalization (#7686)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 09:13:41 +0000 (18:13 +0900)]
[moco-tf] Same padding case for Conv2DBackpropInput Canonicalization (#7686)

* [moco-tf] Same padding case for Conv2DBackpropInput Canonicalization

This commit implements SAME padding case for Conv2DBackpropInput
Canonicalization

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
* Clarify comment

4 years ago[moco-int] Enable test SquaredDifference_001 (#7690)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 08:21:40 +0000 (17:21 +0900)]
[moco-int] Enable test SquaredDifference_001 (#7690)

This will enable SquaredDifference_001 unit test

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years agoIntroduce circle_loader (#7667)
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Mon, 23 Sep 2019 08:12:20 +0000 (11:12 +0300)]
Introduce circle_loader (#7667)

Add `circle_loader` library without implementation.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
4 years ago[moco-tf] Canonicalize rank-0 tensor constants (#7687)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 07:59:37 +0000 (16:59 +0900)]
[moco-tf] Canonicalize rank-0 tensor constants (#7687)

This commit extends ConstCanonicalizer to translate tf.Const nodes that
emits a rank-0 tensor as a canonical.ConstGen node that emits a rank-1
tensor with shape [1].

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years ago[tf2tflite] Enable test SquaredDifference_001 (#7691)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 07:54:30 +0000 (16:54 +0900)]
[tf2tflite] Enable test SquaredDifference_001 (#7691)

This will enable SquaredDifference_001 unit test

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[exo-tflite] Update Reshape exporter (#7680)
남궁석/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 07:35:54 +0000 (16:35 +0900)]
[exo-tflite] Update Reshape exporter (#7680)

* [exo-tflite] Update Reshape exporter

Until now, new shape is exported only by `new_shape` attribute.
However, runtime was receiving new shape by input.
This commit will update `Reshape` exporter to export new shape
not only as attribute but also as input.

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
* add comments

4 years ago[moco-integration-test] Get toolchain from a file (#7678)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 06:07:16 +0000 (15:07 +0900)]
[moco-integration-test] Get toolchain from a file (#7678)

This commit revises runall.sh to reads toolchain configuration from
a file instead of command-line arguments.

This change makes it easy to re-run failed tests for diagnostics.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years agoChange ReduceMean kernel of acl_cl (#7654)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 05:48:48 +0000 (14:48 +0900)]
Change ReduceMean kernel of acl_cl (#7654)

This commit changes ReduceMean kernel of acl_cl.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years agoMake to support EmbeddingLookup op for acl neon (#7663)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 05:32:15 +0000 (14:32 +0900)]
Make to support EmbeddingLookup op for acl neon (#7663)

This commit makes to support EmbeddingLookup op for acl neon.
  - Introduce NEEmbeddingLookup NEEmbeddingLookupKernel
  - Apply NEEmbeddingLookup layer for neurun

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[moco-tf] Enable ResolveSquaredDifference knob (#7670)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 05:27:20 +0000 (14:27 +0900)]
[moco-tf] Enable ResolveSquaredDifference knob (#7670)

This will enable ResolveSquaredDifference knob that decomposes SquaredDifference to Sub and Mul nodes

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[tf2tflite] Enable test Div/Sub 001 (#7673)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 05:11:05 +0000 (14:11 +0900)]
[tf2tflite] Enable test Div/Sub 001 (#7673)

This will enable unit test of Div/Sub 001 that tests broadcasting

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[tf2tflite] Enable UNIT_Squeeze_003 test (#7679)
남궁석/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 04:59:18 +0000 (13:59 +0900)]
[tf2tflite] Enable UNIT_Squeeze_003 test (#7679)

This commit will enable `UNIT_Squeeze_003` test in tf2tflite

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
4 years agoAdd manifest to TensorFlow NET_0028 test (#7672)
박종현/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 04:59:03 +0000 (13:59 +0900)]
Add manifest to TensorFlow NET_0028 test (#7672)

This commit adds a manifest for NET_0028 test which includes a
brief summary.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
4 years agoIntroduce `SupportPermutation` to check if backend supports permutation (#7616)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 04:54:25 +0000 (13:54 +0900)]
Introduce `SupportPermutation` to check if backend supports permutation (#7616)

This commit introduces `SupportPermutation` to check if backend supports permutation.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[nnfw-cmake] Change nnfw variable name used in modules (#7671)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 04:32:51 +0000 (13:32 +0900)]
[nnfw-cmake] Change nnfw variable name used in modules (#7671)

Introduce NNFW_PROJECT_SOURCE_DIR and NNFW_EXTERNALS_DIR to alternate NNFW_XXX variable

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[moco-integration-test] Enable UNIT_Mean_00x tests (#7656)
남궁석/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 04:32:46 +0000 (13:32 +0900)]
[moco-integration-test] Enable UNIT_Mean_00x tests (#7656)

This commit will enable `UNIT_Mean_000` ~ `UNIT_Mean_003`
tests in moco-integration-test

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
4 years ago[tf2tflite] Enable UNIT_Mean_00x tests (#7657)
남궁석/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 04:32:26 +0000 (13:32 +0900)]
[tf2tflite] Enable UNIT_Mean_00x tests (#7657)

This commit will enable `UNIT_Mean_000` ~ `UNIT_Mean_003`
tests in tf2tflite

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
4 years ago[moco_int] Enable Add 005 unit test (#7674)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 04:30:51 +0000 (13:30 +0900)]
[moco_int] Enable Add 005 unit test (#7674)

This will enable Add 005 unit test

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[exo-tflite] Enable Div,Sub conversion knob (#7677)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 04:27:28 +0000 (13:27 +0900)]
[exo-tflite] Enable Div,Sub conversion knob (#7677)

This will enable knob for conversion of Div and Sub

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[neurun] Implement NNAPI get output dimensions (#7653)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 04:26:29 +0000 (13:26 +0900)]
[neurun] Implement NNAPI get output dimensions (#7653)

Inplement NNAPI ANeuralNetworksExecution_getOutputOperandDimensions and wrapper function

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoMake to support DepthToSpace op for acl neon (#7629)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Mon, 23 Sep 2019 02:05:48 +0000 (11:05 +0900)]
Make to support DepthToSpace op for acl neon (#7629)

This commit makes to support DepthToSpace op for acl neon.
  - Introduce NEDepthToSpaceLayerEx and NEDepthToSpaceLayerKernelEx
  - Apply NEDepthToSpaceLayerEx layer for neurun

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[exo-tflite] Enable Div,Sub conversion (#7669)
박세희/On-Device Lab(SR)/Principal Engineer/삼성전자 [Mon, 23 Sep 2019 01:32:23 +0000 (10:32 +0900)]
[exo-tflite] Enable Div,Sub conversion (#7669)

This will enable Eltwise Div/Sub node conversion to TFL node by knob

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
4 years ago[nncc-cmake] Change nncc variable name used in modules (#7662)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Mon, 23 Sep 2019 00:49:35 +0000 (09:49 +0900)]
[nncc-cmake] Change nncc variable name used in modules (#7662)

Introduce NNAS_PROJECT_SOURCE_DIR and NNAS_EXTERNALS_DIR to alternate NNCC_XXX variable
Use workaround for NNAS_EXTERNALS_DIR - CI issue

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[moco-tf] Padding inference for 'valid' padding case (#7650)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Sun, 22 Sep 2019 23:55:41 +0000 (08:55 +0900)]
[moco-tf] Padding inference for 'valid' padding case (#7650)

* [moco-tf] Padding inference for 'valid' padding case

This commit implements padding inference for Conv2DBackpropInput
canonicalization, in case of 'valid' padding.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
* bool ready()

4 years ago[moco-tf] Remove unused function arg (#7607)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Sun, 22 Sep 2019 22:19:17 +0000 (07:19 +0900)]
[moco-tf] Remove unused function arg (#7607)

This commit removes unused function argument in Conv2D canonicalizer.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[cmake] Rename nncc_include to nnas_include (#7659)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 08:34:25 +0000 (17:34 +0900)]
[cmake] Rename nncc_include to nnas_include (#7659)

Rename nncc_include to nnas_include to share modules with nnfw

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[exo-tflite] pass PhaseStrategy from Saturate to Restart (#7652)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Fri, 20 Sep 2019 08:31:43 +0000 (17:31 +0900)]
[exo-tflite] pass PhaseStrategy from Saturate to Restart (#7652)

pass strategy was Restart but changed to Saturate by accident. This commit changes it to Restart again.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[locomotiv] Introduce TensorReduce Operation (#7643)
남궁석/On-Device Lab(SR)/Engineer/삼성전자 [Fri, 20 Sep 2019 07:43:15 +0000 (16:43 +0900)]
[locomotiv] Introduce TensorReduce Operation (#7643)

* [locomotiv] Introduce TensorReduce Operation

This commit will introduce `TensorReduce` operation in `locomotiv`
For now, only supports `ReduceFunc::Mean`

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
* restructure

4 years ago[moco-tf] Register Conv2DBackpropInput canonicalization (#7649)
박천교/On-Device Lab(SR)/Engineer/삼성전자 [Fri, 20 Sep 2019 07:27:16 +0000 (16:27 +0900)]
[moco-tf] Register Conv2DBackpropInput canonicalization (#7649)

This commit registers Conv2DBackpropInput canonicalization.

Signed-off-by: Cheongyo Bahk <ch.bahk@samsung.com>
4 years ago[cmake] Move cmake modules (#7651)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 07:13:34 +0000 (16:13 +0900)]
[cmake] Move cmake modules (#7651)

Move nncc's cmake modules to infra/cmake/modules to share with nnfw

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years ago[exo-tflite] converting and exporting TFLConst (#7638)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Fri, 20 Sep 2019 07:12:22 +0000 (16:12 +0900)]
[exo-tflite] converting and exporting TFLConst (#7638)

With this, TFLConverter start work and TensorExporter exports tensor using TFLConst when the knob is on.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years agoMake to support ReduceMax/ReduceMin op for acl neon (#7599)
장지섭/On-Device Lab(SR)/Engineer/삼성전자 [Fri, 20 Sep 2019 07:10:11 +0000 (16:10 +0900)]
Make to support ReduceMax/ReduceMin op for acl neon (#7599)

This commit makes to support ReduceMax/ReduceMin op for acl neon
  - Introduce NEReductionOperationKernelEx
  - Introduce NEReductionOperationEx and NEReduceOperation
  - Apply NEReduceOperation layer for neurun to enable ReduceMax/ReduceMin op

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
4 years ago[exo-tflite] change TFL node opcode (NOP_CONSTGEN -> CONST) (#7621)
윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자 [Fri, 20 Sep 2019 07:09:17 +0000 (16:09 +0900)]
[exo-tflite] change TFL node opcode (NOP_CONSTGEN -> CONST) (#7621)

Since use TFLConst (not anything like Nop), I changed the opcode also.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
4 years ago[neurun] Implement a hash functor for enum class (#7647)
이한종/On-Device Lab(SR)/Engineer/삼성전자 [Fri, 20 Sep 2019 06:57:25 +0000 (15:57 +0900)]
[neurun] Implement a hash functor for enum class (#7647)

Implement a hash functor for enum class `NodeName` since Android NDK
clang requires it.

* This is a fix for Android Build

Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
4 years ago[nnas-cmd] Use pylint command anywhere (#7648)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 06:56:57 +0000 (15:56 +0900)]
[nnas-cmd] Use pylint command anywhere (#7648)

Allow use pylint command any path under nnfw
- Use "git ls-files" to find '.FORMATDENY' file
- Use pushd/popd to search python files on nnfw's root path

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoUpdate tflite_benchmark_model for TF 1.14 (#7645)
이상규/On-Device Lab(SR)/Principal Engineer/삼성전자 [Fri, 20 Sep 2019 05:27:19 +0000 (14:27 +0900)]
Update tflite_benchmark_model for TF 1.14 (#7645)

It fixes `tflite_benchmark_model` build error that happends after
tensorflow is updated to 1.14.

Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
4 years ago[nncc-cmake] Check external module path existance (#7646)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 04:37:26 +0000 (13:37 +0900)]
[nncc-cmake] Check external module path existance (#7646)

Check external module path existance and download if that is not exists
Same with nnfw

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
4 years agoMake Compatible Shapes for CPU Add (#6540)
이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 04:04:58 +0000 (13:04 +0900)]
Make Compatible Shapes for CPU Add (#6540)

- This commit addes routine for making compatible shapes for CPU Add
  kernel

Signed-off-by: Chunseok Lee <chunseok.lee@samsung.com>
4 years agoImplement NNAPI get output rank function (#7581)
오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 [Fri, 20 Sep 2019 03:42:00 +0000 (12:42 +0900)]
Implement NNAPI get output rank function (#7581)

Inplement NNAPI ANeuralNetworksExecution_getOutputOperandRank and wrapper function
Introduce private field and method in Execution class to check execution is finished

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>