platform/core/ml/nnfw.git
5 years ago[nnc] Add support for freestanding activation functions in tflite importer (#2385)
Vladimir Plazun/AI Tools Lab /SRR/Engineer/삼성전자 [Fri, 23 Nov 2018 17:53:52 +0000 (20:53 +0300)]
[nnc] Add support for freestanding activation functions in tflite importer  (#2385)

While tflite have fused activation functions, some networks use activation functions as separate operations

Signed-off-by: Vladimir Plazun <v.plazun@partner.samsung.com>
5 years ago[tflchef/reverse] Enable INT32 type (#2393)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 08:01:07 +0000 (17:01 +0900)]
[tflchef/reverse] Enable INT32 type (#2393)

This will enable tflchef-reverse to handle INT32 type
- Needed to handle Reshape operator from MobileNet

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Update depthwiseconv2d recipe (use relu6 activation) (#2322)
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 04:07:20 +0000 (13:07 +0900)]
[tflchef] Update depthwiseconv2d recipe (use relu6 activation) (#2322)

depthwiseconv2d recipe will use relu6 activation.
- Mobilenet uses relu6 activation.
- It could be used as a test for relu6 activation.

Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
5 years ago[tflchef] Support int32 tensortype (#2388)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 04:06:01 +0000 (13:06 +0900)]
[tflchef] Support int32 tensortype (#2388)

This will fix to support int32 tensortype with introducing as_tflite_tensortype()

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] enable ReLU6 conversion (#2381)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 01:37:19 +0000 (10:37 +0900)]
[tflchef] enable ReLU6 conversion (#2381)

This will enable ReLU6 in as_tflchef_activation()

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/frontend] Empty DepthwiseConv2D (#2387)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 01:36:42 +0000 (10:36 +0900)]
[enco/frontend] Empty DepthwiseConv2D (#2387)

This will add placeholder for DepthwiseConv2D in frontend
- to check loading mobilenet slim model

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/tfl/frontend] ReLU6 Activation Layer (#2359)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Fri, 23 Nov 2018 01:04:38 +0000 (10:04 +0900)]
[enco/tfl/frontend] ReLU6 Activation Layer (#2359)

* [enco/tfl/frontend] ReLU6 Activation Layer

This commit is for ReLU6 Activation Layer.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* Pr fix: add code to op/Activation.cpp

5 years ago[nnc] Support Pad operation on TFLite Importer and Interpreter (#2330)
Павел Ильютченко/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Thu, 22 Nov 2018 11:30:58 +0000 (14:30 +0300)]
[nnc] Support Pad operation on TFLite Importer and Interpreter (#2330)

* Support Pad op in TFLite imported and interpreter

Signed-off-by: Pavel Iliutchenko <p.iliutchenk@partner.samsung.com>
5 years ago[tflchef] Add DepthwiseConv2D for reverse (#2373)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 08:20:04 +0000 (17:20 +0900)]
[tflchef] Add DepthwiseConv2D for reverse (#2373)

This will add DepthwiseConv2D Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add Conv2D for reverse (#2375)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 07:59:01 +0000 (16:59 +0900)]
[tflchef] Add Conv2D for reverse (#2375)

* [tflchef] Add Conv2D for reverse

This will add Conv2D Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* add comment

5 years ago[tflchef] Add MaxPool2D for reverse (#2376)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 07:48:01 +0000 (16:48 +0900)]
[tflchef] Add MaxPool2D for reverse (#2376)

This will add MaxPool2D Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add ReLU6 for reverse (#2374)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 07:46:43 +0000 (16:46 +0900)]
[tflchef] Add ReLU6 for reverse (#2374)

This will add ReLU6 Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add ReLU for reverse (#2377)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 05:58:55 +0000 (14:58 +0900)]
[tflchef] Add ReLU for reverse (#2377)

This will add ReLU Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add Concat for reverse (#2370)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 05:19:21 +0000 (14:19 +0900)]
[tflchef] Add Concat for reverse (#2370)

This will add Concatenation Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add Reshape for reverse (#2371)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 05:12:23 +0000 (14:12 +0900)]
[tflchef] Add Reshape for reverse (#2371)

This will add Reshape Op for tflchef-reverse

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco] Splitting Frontend.cpp into Op/*.cpp and more (#2362)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 05:08:03 +0000 (14:08 +0900)]
[enco] Splitting Frontend.cpp into Op/*.cpp and more (#2362)

* [enco] Splitting Frontend.cpp into Op/*.cpp and more

Through this commit, Frontend.cpp will be refactored into several files.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* pr fix: remove useless "tflite::"

* pr fix: Split h file into h and cpp file

* pr fix: remove wrong test files

5 years ago[tflchef] Fix FlatBufferIntArrayToVector (#2369)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 01:33:51 +0000 (10:33 +0900)]
[tflchef] Fix FlatBufferIntArrayToVector (#2369)

* [tflchef] Fix FlatBufferIntArrayToVector

This will rename FlatBufferIntArrayToVector to as_index_vector like enco/frontend and correct type of caller
- incorrect type may cause a problem in reading minus values in shape info

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* fix to typename

5 years ago[tflchef] Prepare adding Ops with AveragePool2D (#2368)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 22 Nov 2018 01:33:22 +0000 (10:33 +0900)]
[tflchef] Prepare adding Ops with AveragePool2D (#2368)

This will add AveragePool2D operator handler and needed changes to test.
Also place holders for other operators to prevent merge conflict.

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] generate_recipe for each operators (#2361)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 22:18:40 +0000 (07:18 +0900)]
[tflchef] generate_recipe for each operators (#2361)

This will fill handlers for each operators
- forward filler settings for the operator
- tflchef operator build and I/O settings for corresponding tflite operator

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[nnc] Fix deconv unittest build (#2366)
Efimov Alexander/AI Tools Lab/./Samsung Electronics [Wed, 21 Nov 2018 19:10:05 +0000 (22:10 +0300)]
[nnc] Fix deconv unittest build (#2366)

Fix deconv unit test build and adjust float epsilon

Signed-off-by: Efimov Alexander <a.efimov@samsung.com>
5 years ago[nnc][utils][model_generator] Add converter from tree to tflite. (#2230)
Иван Иванович Кулагин/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 21 Nov 2018 18:19:24 +0000 (21:19 +0300)]
[nnc][utils][model_generator] Add converter from tree to tflite. (#2230)

Implemented converter from randomly generated Tree representation to TFLite model.

Signed-off-by: i-kulagin i.kulagin@samsung.com
5 years ago[nnc] Initial support of various operations in ONNX frontend (#2286)
Андрей Тищенко/AI Tools Lab /SRR/Staff Engineer/삼성전자 [Wed, 21 Nov 2018 18:04:58 +0000 (21:04 +0300)]
[nnc] Initial support of various operations in ONNX frontend (#2286)

Operations supported in ONNX importer:
- Conv2d
- BatchNorm
- Pool
- scale
- dropout
- Elementwise operations (add, sum, mul, max)
- Gemm

Signed-off-by: Andrew V. Tischenko <a.tischenko@partner.samsung.com>
5 years ago[nnc] Conv transpose in SoftBackend (#2290)
Андрей Шедько/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Wed, 21 Nov 2018 17:42:29 +0000 (20:42 +0300)]
[nnc] Conv transpose in SoftBackend (#2290)

Implements Deconvolution in Soft Backend, caffe style transfer model is fully supported.

Signed-off-by: Andrei Shedko a.shedko@partner.samsung.com
5 years ago[nnc] Add `shape` parameter to constructors of operations that cannot infer shapes...
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 21 Nov 2018 16:23:36 +0000 (19:23 +0300)]
[nnc] Add `shape` parameter to constructors of operations that cannot infer shapes of their outputs (#2338)

* Add `shape` parameter to constructor of `VariableOp`;
* Add `shape` parameter to constructor of `ReshapeOp`.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
5 years ago[tflchef] Implement generate_recipe (#2360)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 10:31:57 +0000 (19:31 +0900)]
[tflchef] Implement generate_recipe (#2360)

* [tflchef] Implement generate_recipe

This implements partial of generate_recipe() for operators, operands and inputs,outputs.

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* remove unused

* reorder a line

* fix typo

* need check quantization

* fix typo

5 years ago[tflchef] Introduce TFliteOpChef and Registry (#2357)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 10:25:15 +0000 (19:25 +0900)]
[tflchef] Introduce TFliteOpChef and Registry (#2357)

This will introduce TFliteOpChef base class for Operator handlers for translation and
TFliteOpRegistry for registeration of handlers.
All headers are included in RecipeChef to check compilation and as it will be used later codes.

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Adds "comment" into recipe (#2356)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 09:51:18 +0000 (18:51 +0900)]
[tflchef] Adds "comment" into recipe (#2356)

Now, recipe file can have "comment" field, which is ignored by tflchef.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[tflchef] Introduce TFliteImport (#2355)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 08:15:16 +0000 (17:15 +0900)]
[tflchef] Introduce TFliteImport (#2355)

* [tflchef] Introduce TFliteImport

This will introduce TFliteImport class that provides needed information from tflite::Model to build recipe

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* add brief

5 years ago[enco/tfl/frontend] Modified wrong test number (#2354)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 05:11:59 +0000 (14:11 +0900)]
[enco/tfl/frontend] Modified wrong test number (#2354)

Conv_003 exists without Conv_002. So Conv_003 is renamed to Conv_002.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[tflchef] Introduce convert helpers (#2351)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 05:11:09 +0000 (14:11 +0900)]
[tflchef] Introduce convert helpers (#2351)

This will introduce helper functions to convert tflite types to tflchef types

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Store ModelRecipe to file (#2352)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 05:10:39 +0000 (14:10 +0900)]
[tflchef] Store ModelRecipe to file (#2352)

This will implement storing ModelRecipe to a file

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/tfl/frontend] Bug fix on padding (#2353)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 05:10:19 +0000 (14:10 +0900)]
[enco/tfl/frontend] Bug fix on padding (#2353)

IFM is NHWC format but the padding code was written assuming that IFM is NCHW.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[enco/tfl/frontend] Refactoring RELU activation (#2313)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 04:44:45 +0000 (13:44 +0900)]
[enco/tfl/frontend] Refactoring RELU activation (#2313)

* [enco/tfl/frontend] Refactoring RELU activation

RELU activation is commonly used by many operations. So this activation is extracted as a separate method.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* returning ifm

* ofm as param

* remove empty line

* pr fix: remove output param

* pr fix : comment, renaming vars, etc.

5 years ago[tflchef] Call tflchef from Driver (#2350)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 02:51:21 +0000 (11:51 +0900)]
[tflchef] Call tflchef from Driver (#2350)

This will make Driver call tflchef tflite load and save functions

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] RawModel loader (#2347)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 02:18:49 +0000 (11:18 +0900)]
[tflchef] RawModel loader (#2347)

This will introduce TFlite raw model loader
- copied from enco/frontend

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Introduce entry points for tflchef-reverse (#2345)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 02:18:20 +0000 (11:18 +0900)]
[tflchef] Introduce entry points for tflchef-reverse (#2345)

* [tflchef] Introduce entry points for tflchef-reverse

This will introduce two empty functions for tflchef-reverse tool
- CMakeLists.txt file to build a library
- This library will read TF lite file and drop tflchef recipe file

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* change header ifdef name

5 years ago[tflchef] Introduce reverse tool driver (#2346)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 21 Nov 2018 01:19:12 +0000 (10:19 +0900)]
[tflchef] Introduce reverse tool driver (#2346)

[tflchef] Introduce reverse tool driver

5 years ago[nnc] fix static analysis warning and add missed break (#2287)
Павел Фаттахов/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Tue, 20 Nov 2018 18:10:31 +0000 (21:10 +0300)]
[nnc] fix static analysis warning and add missed break (#2287)

- Add assert, which protects us from passing a null pointer to a function.
- Add break statement into switch in tflite_importer.

Signed-off-by: Pavel <p.fattakhov@partner.samsung.com>
5 years ago[tflchef] Add Conv2D test recipe (#2340)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 20 Nov 2018 09:52:43 +0000 (18:52 +0900)]
[tflchef] Add Conv2D test recipe (#2340)

This will add Conv2D test recipe for tflchef.
- there was existing readme to show readme example works but there was no conv2d test

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[nnc] Make constructors of `Operations` accept other `Operation`s as arguments (...
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Mon, 19 Nov 2018 15:32:09 +0000 (18:32 +0300)]
[nnc] Make constructors of `Operations` accept other `Operation`s as arguments (#2292)

* Make constructors of `Operations` accept `Operation`s they depend on.
* Remove `connectInput` method, which became redundant.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
5 years ago[nnc] split model runner script into 4 scripts (#2285)
Павел Фаттахов/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Mon, 19 Nov 2018 14:19:58 +0000 (17:19 +0300)]
[nnc] split model runner script into 4 scripts (#2285)

split model runner script into 4 scripts

Signed-off-by: Pavel <p.fattakhov@partner.samsung.com>
5 years ago[enco] Fix typo (#2334)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Mon, 19 Nov 2018 08:17:54 +0000 (17:17 +0900)]
[enco] Fix typo (#2334)

This will fix typo in OwnedFileDescriptor comment

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Add ReLU6 in activation (#2312)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Fri, 16 Nov 2018 02:06:49 +0000 (11:06 +0900)]
[tflchef] Add ReLU6 in activation (#2312)

This commit will add ReLU6 in activation

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[enco] Remove unused variable (#2311)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 15 Nov 2018 23:09:30 +0000 (08:09 +0900)]
[enco] Remove unused variable (#2311)

This commit will remove unused varuable

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[nnkit] Link backend_support library with dl (#2310)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 23:08:48 +0000 (08:08 +0900)]
[nnkit] Link backend_support library with dl (#2310)

Backend support library uses dlopen internally, but is not linked with
dl currently.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco] Set subnet output correctly on branching (#2309)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 23:07:52 +0000 (08:07 +0900)]
[enco] Set subnet output correctly on branching (#2309)

The current implementation specifies a bag as a subnet output if the subnet
accesses it and this bag is accessed outside of this subnet.

It turns out that this algorithm may specify a network input as a subnet output
as it does not consider this bag is updated by this subnet.

This commit fixes this bug and adds a related test.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[tflchef] Use safemain (#2305)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 23:07:05 +0000 (08:07 +0900)]
[tflchef] Use safemain (#2305)

Now, tflchef and tflchef-file uses the main function that safemain
provides.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[nnc] Implemented caffe2 importer interface (#2233)
Ivan Vagin/AI Tools Lab /SRR/Engineer/삼성전자 [Thu, 15 Nov 2018 16:16:31 +0000 (19:16 +0300)]
[nnc] Implemented caffe2 importer interface (#2233)

* Implemented caffe2 importer interface
* Minor fixes in driver sanity checks
* Inited caffe2 model dumper

Signed-off-by: Ivan Vagin <ivan.vagin@samsung.com>
5 years ago[nnc] Remove redundant code (#2314)
Efimov Alexander/AI Tools Lab/./Samsung Electronics [Thu, 15 Nov 2018 16:02:30 +0000 (19:02 +0300)]
[nnc] Remove redundant code (#2314)

Remove redundant code from c++ snippet that prevents artifact build

Signed-off-by: Efimov Alexander <a.efimov@samsung.com>
5 years ago[nnc] Replace layout of operations in cpu backend (#2258)
Efimov Alexander/AI Tools Lab/./Samsung Electronics [Thu, 15 Nov 2018 13:51:10 +0000 (16:51 +0300)]
[nnc] Replace layout of operations in cpu backend (#2258)

Order of operations in artifact is changed to RPO to optimize usage of memory caches

Signed-off-by: Efimov Alexander <a.efimov@samsung.com>
5 years ago[nnc] Tfl importer eltwise (#2208)
Андрей Шедько/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Thu, 15 Nov 2018 13:04:26 +0000 (16:04 +0300)]
[nnc] Tfl importer eltwise (#2208)

Added elementwise operations, Tanh and Transposed Conv to tfLite importer

Signed-off-by: Andrei Shedko <a.shedko@partner.samsung.com>
5 years ago[coco] Deprecate FeatureObject create with Shape (#2302)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 06:35:02 +0000 (15:35 +0900)]
[coco] Deprecate FeatureObject create with Shape (#2302)

This commit removes create method that takes a FeatureShape from
ObjectManager class and revises related tests.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco/frontend] Relu activation layer for Conv2D (#2306)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 15 Nov 2018 06:33:14 +0000 (15:33 +0900)]
[enco/frontend] Relu activation layer for Conv2D (#2306)

* [enco/frontend] Relu activation layer for Conv2D

This commit adds Relu activation layer for Conv2D and test.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* Check NONE for actuvation

5 years ago[enco] Introduce enco_dump_all_objects (#2304)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 04:01:34 +0000 (13:01 +0900)]
[enco] Introduce enco_dump_all_objects (#2304)

With this debugging helper, users easily enumerates all the objects in a
module with interactive debugger.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco.tflite] Remove deprecated create method use (#2303)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 03:59:54 +0000 (12:59 +0900)]
[enco.tflite] Remove deprecated create method use (#2303)

FeatureObject create method (with a FeatureShape) in ObjectManager is
deprecated.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco] Fix typos in some comments (#2301)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 15 Nov 2018 02:35:31 +0000 (11:35 +0900)]
[enco] Fix typos in some comments (#2301)

There are same typos in some comments
This commit will fix it

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years agoIntroduce 'safemain' library (#2300)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 02:25:53 +0000 (11:25 +0900)]
Introduce 'safemain' library (#2300)

This commit introduces 'safemain' library which provides 'main' function
that captures std::exception on RELEASE build.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco] Modify some misunderstandable parameters name (#2297)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 15 Nov 2018 02:23:23 +0000 (11:23 +0900)]
[enco] Modify some misunderstandable parameters name (#2297)

Some parameters are started with underbar, which can be misunderstood as private member variable
This commit will fix them not to be misunderstood

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[enco] Extend frontend for ReLU (#2298)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 15 Nov 2018 02:03:40 +0000 (11:03 +0900)]
[enco] Extend frontend for ReLU (#2298)

This commit will extend the frontend of enco for ReLU

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[enco] Introduce enco_dump_all_bags (#2293)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 15 Nov 2018 01:50:32 +0000 (10:50 +0900)]
[enco] Introduce enco_dump_all_bags (#2293)

This commit introduces enco_dump_all_bags which shows the details of all
the allocated bags in a given module.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco/frontend] Support Reshape of TF lite (#2294)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Thu, 15 Nov 2018 01:48:47 +0000 (10:48 +0900)]
[enco/frontend] Support Reshape of TF lite (#2294)

* [enco/frontend] Support Reshape of TF lite

This will add Reshape operator for TF lite import

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* use ifm_bag->size()

* change test out shape

* fix shape again

* fix typo

5 years ago[enco] Extend frontend for ReLU6 (#2275)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 15 Nov 2018 00:29:02 +0000 (09:29 +0900)]
[enco] Extend frontend for ReLU6 (#2275)

* [enco] Extend frontend for ReLU6

This commit will extend the frontend of enco for ReLU6

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

5 years ago[enco/frontend] test case for MaxPool2D tflite frontend (#2272)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 23:51:53 +0000 (08:51 +0900)]
[enco/frontend] test case for MaxPool2D tflite frontend (#2272)

This commit adds a test case for MaxPool2D tflite frontend

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[enco/frontend] fix bug about padding calculation (#2280)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 23:51:27 +0000 (08:51 +0900)]
[enco/frontend] fix bug about padding calculation (#2280)

* [enco/frontend] fix bug about padding calculation

Wrong layout (NCHW) of kernel caused this bug. Fixed to NHWC.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* more complex test (no identical number in ifm shape)

5 years ago[enco] Show Conv2D op details (#2281)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 14 Nov 2018 23:51:13 +0000 (08:51 +0900)]
[enco] Show Conv2D op details (#2281)

With this commit, enco_dump_all_ops and enco_dump_op methods show the
additional information if a given op is a Conv2D op.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[tflchef] Support ReLU in tflchef (#2283)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 14 Nov 2018 23:50:53 +0000 (08:50 +0900)]
[tflchef] Support ReLU in tflchef (#2283)

This commit will enable ReLU in tflchef

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[nnc] Replace ASSERT_NEAR with custom comparator (#2231)
Efimov Alexander/AI Tools Lab/./Samsung Electronics [Wed, 14 Nov 2018 18:37:51 +0000 (21:37 +0300)]
[nnc] Replace ASSERT_NEAR with custom comparator (#2231)

Replace ASSERT_NEAR in soft backend unit tests with custom ULP comparator to add flexibility

Signed-off-by: Efimov Alexander <a.efimov@samsung.com>
5 years agoadded assert for result of ftell function (#2278)
Павел Фаттахов/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Wed, 14 Nov 2018 18:28:31 +0000 (21:28 +0300)]
added assert for result of ftell function (#2278)

added assert for result of ftell function in interpreter.

Signed-off-by: Pavel <p.fattakhov@partner.samsung.com>
5 years ago[nnc] Make the class `Graph` intrusive (#2261)
Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자 [Wed, 14 Nov 2018 17:47:36 +0000 (20:47 +0300)]
[nnc] Make the class `Graph` intrusive (#2261)

* Remove `INode`, `AbstractNode`, `Node`, `OpDescription` classes.
* Add `Operation` class, derive all `Op`s from it.
* Modify `Graph` to work with `Operation`s instead of `Node`.

Signed-off-by: Sergei Barannikov <s.barannikov@samsung.com>
5 years ago[enco/frontend] MaxPool2D for tflite frontend (#2270)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 10:45:47 +0000 (19:45 +0900)]
[enco/frontend] MaxPool2D for tflite frontend (#2270)

* [enco/frontend] MaxPool2D for tflite frontend

This commit adds MaxPool2D for tflite frontend.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* 'auto' to explicit type

5 years ago[tflchef] Support DepthwiseConv2D (#2277)
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 10:12:42 +0000 (19:12 +0900)]
[tflchef] Support DepthwiseConv2D (#2277)

This patch extends tflchef for DepthwiseConv2D operator.

Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
5 years ago[enco] Bugfix about ReLU6 for backend (#2273)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 14 Nov 2018 08:19:29 +0000 (17:19 +0900)]
[enco] Bugfix about ReLU6 for backend (#2273)

There are some typos for ReLU6 operation
This commit will fix them

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[tflchef] Support Reshape (#2268)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 08:11:02 +0000 (17:11 +0900)]
[tflchef] Support Reshape (#2268)

This will extend to support Reshape operator

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/frontend] Added test cases for AvgPool2D (#2244)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 04:35:50 +0000 (13:35 +0900)]
[enco/frontend] Added test cases for AvgPool2D (#2244)

* [enco/frontend] added a test case for AvgPool2D

Two test cases for AvgPool2D are added. (paddiang for SAME and VALID)

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* avgpool2d -> averagepool2d

5 years ago[enco] Rename caffe tests (#2267)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 14 Nov 2018 03:58:12 +0000 (12:58 +0900)]
[enco] Rename caffe tests (#2267)

This commit appends layer prefix (such as Convolution) to each caffe tests and
assign new numbers in order to make it easy to find.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco] Extend backend for ReLU6 (#2265)
남궁석/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 14 Nov 2018 03:53:07 +0000 (12:53 +0900)]
[enco] Extend backend for ReLU6 (#2265)

This commit will extend the backend of enco for ReLU6

Signed-off-by: Seok NamKoong <sk.namkoong@samsung.com>
5 years ago[enco/frontend] Test for TF lite concat (#2264)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 03:52:31 +0000 (12:52 +0900)]
[enco/frontend] Test for TF lite concat (#2264)

This will add test recipe for Concat operator

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/frontend] Introduce ConcatenationGraphBuilder (#2262)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 01:54:17 +0000 (10:54 +0900)]
[enco/frontend] Introduce ConcatenationGraphBuilder (#2262)

* [enco/frontend] Introduce ConcatenationGraphBuilder

This will introduce ConcatenationGraphBuilder for importing Concat operator from TF lite

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* remove unused

5 years ago[enco/frontend] Revise Conv2D build using copy (#2255)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 14 Nov 2018 01:11:38 +0000 (10:11 +0900)]
[enco/frontend] Revise Conv2D build using copy (#2255)

* [enco/frontend] Revise Conv2D build using copy

This will revise Conv2D to use InstrBuilder::copy method so that cascading new coco IR operator would be easy

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* renamve var

* fix comment

5 years ago[enco] Remove caffe build tests (#2263)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 14 Nov 2018 01:09:58 +0000 (10:09 +0900)]
[enco] Remove caffe build tests (#2263)

Build test was introduced to check whether the source code that enco emits
is valid in terms of its syntax at the early stage, but this test is no longer
meaningful.

This commit removes caffe build tests and related infrastructure to
simplify build process.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[nnc] Support PAD operation on modelIR (#2232)
Павел Ильютченко/AI Tools Lab /SRR/Assistant Engineer/삼성전자 [Tue, 13 Nov 2018 16:59:29 +0000 (19:59 +0300)]
[nnc] Support PAD operation on modelIR (#2232)

* Create new class for PAD operation and visitors
* Support pad operation in shape Inference
* fix ElementwiseOp dump in irDotDumper

Signed-off-by: Pavel Iliutchenko <p.iliutchenk@partner.samsung.com>
5 years ago[nnc] Problems fixed during the debug of the Convolution and the Softmax operations...
Тимур Отеллович Аблязимов/AI Tools Lab /SRR/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 16:27:02 +0000 (19:27 +0300)]
[nnc] Problems fixed during the debug of the Convolution and the Softmax operations (#2142)

- Tensor and tensor shapes transposed in certain places.
- Serialization operation changed to save tensors starting from lower to higher dimensions (to be easily readable by the standard ACL routine).

Signed-off-by: Timur Ablyazimov <t.ablyazimov@samsung.com>
5 years agoFirst implementation of some Onnx operators. (#2140)
Андрей Тищенко/AI Tools Lab /SRR/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 16:24:40 +0000 (19:24 +0300)]
First implementation of some Onnx operators. (#2140)

The following operators were implemented (without any debugging):
Dropuot, Concat, Scale, Reshape, Rel, Softmax.

Signed-off-by: Andrew V. Tischenko a.tischenko@partner.samsung.com
5 years ago[enco] Introduce enco_dump_all_instrs (#2252)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 10:43:38 +0000 (19:43 +0900)]
[enco] Introduce enco_dump_all_instrs (#2252)

This commit implements enco_dump_all_instrs debugging helper.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco/frontend] Add AvgPool2DGraphBuilder into registry (#2253)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 10:43:11 +0000 (19:43 +0900)]
[enco/frontend] Add AvgPool2DGraphBuilder into registry (#2253)

With this commit, AvgPool2DGraphBuilder is added into registry.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[enco/frontend] Add AvgPool2DGraphBuilder for tflite frontend (#2249)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 10:14:37 +0000 (19:14 +0900)]
[enco/frontend] Add AvgPool2DGraphBuilder for tflite frontend (#2249)

* [enco/frontend] Add AvgPool2DGraphBuilder for tflite frontend

AvgPool2DGraphBuilder was added for tflite frontend.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
* Divisor , assert for fused activation

* tensorflow -> tensorflow lite

5 years ago[enco/frontend] Introduce copy to InstrBuilder (#2251)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 10:14:04 +0000 (19:14 +0900)]
[enco/frontend] Introduce copy to InstrBuilder (#2251)

* [enco/frontend] Introduce copy to InstrBuilder

This will add copy method to InstrBuilder to make a copy instruction

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* add comment

5 years ago[tflchef] Support MaxPool2D (#2248)
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 10:11:50 +0000 (19:11 +0900)]
[tflchef] Support MaxPool2D (#2248)

This patch extends tflchef for MaxPool2D operator.

Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
5 years ago[enco] Common nullptr handler in debuging API (#2247)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 09:40:09 +0000 (18:40 +0900)]
[enco] Common nullptr handler in debuging API (#2247)

With this commit, every debugging API shares the same nullptr handler
(which is generated via macro).

This change makes it easy to extend debugging API.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco] Do NOT eliminate output objects (#2246)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 08:06:00 +0000 (17:06 +0900)]
[enco] Do NOT eliminate output objects (#2246)

The current implementation of identical object reducation pass
may eliminate output objects (i.e. objects over output bags), which
results in incorrect results.

With this commit, identical object reducation pass no longer eliminates
output objects.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco/frontend] Add padding functions for conv and avgpool (#2243)
윤현식/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 06:08:44 +0000 (15:08 +0900)]
[enco/frontend] Add padding functions for conv and avgpool (#2243)

Add two more padding functions: One is for conv2d, another for avgpool.

Signed-off-by: Hyun Sik Yoon <hyunsik.yoon@samsung.com>
5 years ago[enco] remove underscore in build method argument (#2242)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 05:16:31 +0000 (14:16 +0900)]
[enco] remove underscore in build method argument (#2242)

This will remove underscore used in build method argument

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[tflchef] Remove unnecessary bias input operand from avgpool2d (#2241)
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 05:08:16 +0000 (14:08 +0900)]
[tflchef] Remove unnecessary bias input operand from avgpool2d (#2241)

`bias` operand is not used in averagepool2d.
The operand is removed from recipe.

Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
5 years ago[enco/frontend] Enable Conv2D test (#2240)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 04:44:22 +0000 (13:44 +0900)]
[enco/frontend] Enable Conv2D test (#2240)

This will add INFERENCE file to enable TF lite Conv2D operator load test

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/frontend] Add convolution recipe file Conv2D_000 (#2167)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 04:12:18 +0000 (13:12 +0900)]
[enco/frontend] Add convolution recipe file Conv2D_000 (#2167)

This will add tflchef recipe file that has one convolution operator

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco/frontend] Fix shape input for object layout (#2237)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 02:14:42 +0000 (11:14 +0900)]
[enco/frontend] Fix shape input for object layout (#2237)

This will fix shape input for object layout with correct value

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[enco] Add enco_dump_op_tree debugging helper (#2227)
박종현/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 13 Nov 2018 01:53:35 +0000 (10:53 +0900)]
[enco] Add enco_dump_op_tree debugging helper (#2227)

This commit implements enco_dump_op_tree debugging helper which dumps
all the descendents of a given op as a tree.

Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
5 years ago[enco/frontend] revise opcode_name method (#2228)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 01:53:15 +0000 (10:53 +0900)]
[enco/frontend] revise opcode_name method (#2228)

* [enco/frontend] revise opcode_name method

This will revisit opcode_name method with opcode index value to invalid opcode name

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
* return non const

5 years ago[enco/frontend] Remove load function asserts (#2236)
박세희/동작제어Lab(SR)/Principal Engineer/삼성전자 [Tue, 13 Nov 2018 01:49:43 +0000 (10:49 +0900)]
[enco/frontend] Remove load function asserts (#2236)

This will remove model checking asserts to enable Conv2D operation import

Signed-off-by: SaeHie Park <saehie.park@samsung.com>
5 years ago[nnc] Implement group in CLI (#2235)
Роман Михайлович Русяев/AI Tools Lab /SRR/Staff Engineer/삼성전자 [Mon, 12 Nov 2018 21:02:53 +0000 (00:02 +0300)]
[nnc] Implement group in CLI (#2235)

* add group in Option class
* adapt command line parser to work with grouped options
* refactoring of CommandLine accodring to coding style

Signed-off-by: Roman Rusyaev <r.rusyaev@samsung.com>