platform/core/api/mediavision.git
2 years agomv_roi_tracker: update lib dependency
Hyunsoo Park [Fri, 20 May 2022 06:31:48 +0000 (15:31 +0900)]
mv_roi_tracker: update lib dependency

[Version] : 0.21.7-0
[Issue type] : bug fix

Change-Id: Ie04af3548435608c6cf0d485c831b9b730b8378f
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years agomv_machine_learning: Code refactoring to Prepare function
Inki Dae [Wed, 18 May 2022 06:35:39 +0000 (15:35 +0900)]
mv_machine_learning: Code refactoring to Prepare function

[Version] : 0.21.6-0
[Issue type] : code refactoring

Did code refactoring to mv_inference_open.cpp file by moving
some code to proper place and dropping Prepare function.
Only what Prepare function does is to check given target device types
and set them to a given backend engine but when BindBackend functin
is called in Inference class, SetTargetDevice is called.
It's code duplication and no need anymore.

In addition, call missed UnbindBackend call in error case.

Change-Id: I16aa6c79587175de89ea76891c10b8a75fa53de8
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: code refactoring to mv_inference_open
Inki Dae [Tue, 17 May 2022 10:15:19 +0000 (19:15 +0900)]
mv_machine_learning: code refactoring to mv_inference_open

[Version] : 0.21.5-0
[Issue type] : code refactoring

Did code refactoring to mv_inference_open.cpp file by moving
IsTargetDeviceSupported function to Inference class because
this function is valid only after Bind call. So no reason
for mv_inference_open.cpp to call IsTargetDeviceSupport function
to check if a given device type is valid or not.

And configure backend type as soon as getting the backend type from
the default ini file.

Change-Id: I5d51fb7c71fc737d111f0058e2b1c3ddf1ebf75f
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: code refactoring to Face Recognition
Inki Dae [Fri, 13 May 2022 07:39:40 +0000 (16:39 +0900)]
mv_machine_learning: code refactoring to Face Recognition

[Version] : 0.21.4-0
[Issue type] : code refactoring

Did code refactoring to Face Recognition Framework by doing,
- Make a top module of Face Recognition Framework to pass
  backend engine information such as bacnend type and target device type.
- Move binding code of training engine to Initialize function from
  RegisterNewFace and RecognizeFace functions. By doing this,
  it enhance initial launching time because engine binding is done
  only one time.
- Introduce ClearDataSet API which is a internal API of
  TrainingModel class to drop all previous dataset data.
- Rename Prepare function to ImportLabel one because this function
  imports the label data to in-memory from a file only now.

This is just a step for next code refactoring.

Change-Id: If8b4021793e62c290ce59362ed605215c2bf5e3e
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: correct variable types and cleanup
Inki Dae [Fri, 13 May 2022 04:20:20 +0000 (13:20 +0900)]
mv_machine_learning: correct variable types and cleanup

[Version] : 0.21.3-0
[Issue type] : code cleanup

Corrected types of width and height values and dropped commeted code.

Change-Id: Ic94297e97b4ecb420675176a826a8df3febed134
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: drop dead code
Inki Dae [Thu, 12 May 2022 10:03:29 +0000 (19:03 +0900)]
mv_machine_learning: drop dead code

[Version] : 0.21.2-0
[Issue type] : code cleanup

Change-Id: I2939f8f21f798d40c4315ac628355d40d74ed0be
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: pass mv_source_h instead of raw vector
Inki Dae [Thu, 12 May 2022 09:55:51 +0000 (18:55 +0900)]
mv_machine_learning: pass mv_source_h instead of raw vector

[Version] : 0.21.1-0
[Issue type] : code cleanup

Cleaned up face recognition framework by passing mv_source_h
to face recognition framework instead of raw vector.
By doing this, converting mv_source_h to raw vector is done
in face recognition framework internally.

It is just a step for code refactoring to InferenceEngineHelper class.

Change-Id: I898e19bb8b1ba7d0380f45542ac7fbb8d01f5c59
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: drop deprecated API
Inki Dae [Wed, 11 May 2022 09:05:56 +0000 (18:05 +0900)]
mv_machine_learning: drop deprecated API

[Version] 0.21.0-0
[Issue type] code cleanup

Dropped the deprecated API and relevant code.

Change-Id: Id8d18955dd89a27eb56bc929bd79ae67741680de
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agotest/machine_learning: do not verify label value
Inki Dae [Wed, 11 May 2022 05:44:50 +0000 (14:44 +0900)]
test/machine_learning: do not verify label value

[Version] : 0.20.2-0
[Issue type] : fixup

Do not verify label name. Instead, just verify if API works or not,
and print out the accuracy.

Change-Id: I7432c7ee8055e49a8f51571aaf515a62f4747f75
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: make abstract class
Inki Dae [Wed, 11 May 2022 01:46:35 +0000 (10:46 +0900)]
mv_machine_learning: make abstract class

[Version] : 0.20.1-0
[Issue type] : code cleanup

Made two classes, DataSetManager and TrainingModel, as abstract class
by changing virtual functions to pure virtual ones.

Change-Id: I49d1cba4fd8fb96b065f18380f4b109928c1e1fe
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: add tracker apis
Hyunsoo Park [Tue, 26 Apr 2022 05:32:54 +0000 (14:32 +0900)]
mv_machine_learning: add tracker apis

[Version] : 0.20.0-0
[Issue type] : new feature

Tracking apis is track roi image are released.
Currently it is KCF algorithm based.

[CAPI Description]
    mv_tracker_create
    - Create tracker handle and set desired initial roi to be tracked.
    mv_tracker_perform
- Return roi value in given frame based on initial roi.
  It also return track result whether tracking is sucessful or not.
    mv_tracker_destroy
    - Destroy a given handle.

Change-Id: I8bdfd6d240892b843cf80acdc6aaa338736170b6
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years agomv_machine_learning: revise FeatureVectorManager class
Seungbae Shin [Fri, 6 May 2022 10:26:10 +0000 (19:26 +0900)]
mv_machine_learning: revise FeatureVectorManager class

[Version] 0.19.4-0
[Issue type] refactoring

- Remove unnecessary include from header files
- Make FeatureVectorManager as a abstract class
- Renaming for file stream object
- Use an initialization constructor for fstream
- Apply some useful keyword to help readability
- Remove unnecessary fstream manual closing due to RAII

Change-Id: I0793fb1fd186a3a11bf3fe67b7806e1b8b3ca3da

2 years agoFix SVACE defects
Seungbae Shin [Fri, 6 May 2022 05:13:28 +0000 (14:13 +0900)]
Fix SVACE defects

[Version] 0.19.3-0
[Issue type] Svace

- SIGNED_TO_BIGGER_UNSIGNED
- INTEGER_OVERFLOW

Change-Id: I0bbc17a0a789071b226a1d3d843fdb540bf4686a

2 years agomv_machine_learning: refactoring ml exception
Seungbae Shin [Wed, 27 Apr 2022 05:49:04 +0000 (14:49 +0900)]
mv_machine_learning: refactoring ml exception

[Version] 0.19.2-0
[Issue type] refactoring

- Add BaseException class, an abstract class to handle error code.
- Remove default parameter if possible which is not that useful.
- Remove unnecessary catch section if just propagating is performed.
- Catch by const reference.

Change-Id: Ia720ccbd6dfc8a0ad46d600f62534026df3a7d88

2 years agomv_machine_learning: Fix UNINIT.CTOR.MANY defect on testsuite
Seungbae Shin [Thu, 21 Apr 2022 01:29:23 +0000 (10:29 +0900)]
mv_machine_learning: Fix UNINIT.CTOR.MANY defect on testsuite

[Version] : 0.19.1-0
[Issue type] Svace

Change-Id: I47d5a6a5e70724aff5d49d51f0ee2901105047e6

2 years agomv_machine_learning: add accuracy measurement app for face recognition
Inki Dae [Wed, 20 Apr 2022 05:39:44 +0000 (14:39 +0900)]
mv_machine_learning: add accuracy measurement app for face recognition

[Version] : 0.19.0-0
[Issue type] : new feature

Added an application for measuring the accuracy of face recognition.

Change-Id: I0d41f9619b09dd93b544a77cce6d187b3f7a7723
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: tune face recognition
Inki Dae [Fri, 15 Apr 2022 01:50:39 +0000 (10:50 +0900)]
mv_machine_learning: tune face recognition

[Version] : 0.18.4-0
[Issue type] enhancement

Tuned face recognition accuracy by dropping flip data augmentation and
chaning decision threshold value.

Change-Id: I505628a1d64ff4460eb2760bc9cedb40fd7e65a8
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: handle exception case correctly
Inki Dae [Fri, 15 Apr 2022 01:45:51 +0000 (10:45 +0900)]
mv_machine_learning: handle exception case correctly

[Version] 0.18.3-0
[Issue type] : bug fix

Handled an exception case correctly. In case that internal model file
doesn't exist, this patch returns MEDIA_VISION_ERROR_INVALID_PATH
so that app can check such exception case correctly.

Change-Id: I2494ebaf69931c470ff1a7b82e581745030a0edd
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: call prepare at inference request
Inki Dae [Fri, 8 Apr 2022 06:12:17 +0000 (15:12 +0900)]
mv_machine_learning: call prepare at inference request

[Version] : 0.18.2-0
[Issue type] : bug fix

Called Prepare function at inference request because
training model file can be updated in runtime so
every time inference is requested model and relevant
data should be loaded again.

Change-Id: If3c39d24ea7121262818a811ce433ffe505b94df
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoRevert yolov5 test cases
Tae-Young Chung [Tue, 5 Apr 2022 03:11:06 +0000 (12:11 +0900)]
Revert yolov5 test cases

[Version] 0.18.1-0
[Issue type] revert

After redesigning test cases of yolov5 test cases, it will be back.
Now, revert them.

Change-Id: I475a74448fd011f33d162d7ba3974a0ce01baa3c
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agomv_machine_learning: code refactoring to face recognition
Inki Dae [Wed, 30 Mar 2022 10:03:40 +0000 (19:03 +0900)]
mv_machine_learning: code refactoring to face recognition

[Version] : 0.18.1-0
[Issue type] : code refactoring

Did code refactoring to face recognition framework.
What this refactroing does
- Use unique_ptr instead of shared_ptr by moving data augmentation feature
  dependency to FaceRecognition class.
- Drop duplicated resize operatioon of data augmentation.
- Drop mv_face_recognition_configure function not needed anymore.

Change-Id: I2aceabf82284497f84abe8395d3805315e15185e

2 years agomv_machine_learning: add data augmentation feature
Inki Dae [Tue, 29 Mar 2022 03:51:08 +0000 (12:51 +0900)]
mv_machine_learning: add data augmentation feature

[Version] 0.18.0-0
[Issue type] new feature

Added data augmentation feature which generates several input images
preprocessed in various ways to extend a given input data to mutiple
ones. With this feature, the accuracy is a little improved.
So this patch also corrects a wrong expected answer of test case.

As of now, this feature supports two preprocessing classes - horizontal
flip and rotation, and it uses default and flip classes in default.

Change-Id: I0e18e761c020ffaa8e4cca660f631413c3d5c69c
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: make sure to release previous data
Inki Dae [Tue, 22 Mar 2022 12:15:08 +0000 (21:15 +0900)]
mv_machine_learning: make sure to release previous data

[Version] : 0.17.3-0
[Issue type] : bug fix

mv_face_recognition_prepare function can be called several times
after mv_face_recognition_create function is called one time.
So make sure to drop previous data every time mv_face_recognition_prepare
function is called.

Change-Id: I9626c959d80f981562c4666b64503834ccdbfbbf
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoadd library dependency to training library
Inki Dae [Tue, 22 Mar 2022 09:53:42 +0000 (18:53 +0900)]
add library dependency to training library

[Version] : 0.17.2-0
[Issue type] : bug fix

Added opencv_imgcodecs library dependency to training library.

Change-Id: Ia7da6e2271051ac5fc283661e59d59c2c4202a59
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoupdate pc files
Inki Dae [Mon, 21 Mar 2022 10:31:14 +0000 (19:31 +0900)]
update pc files

[Version] : 0.17.1-0
[Issue type] : bug fix

Updated pc files for training and face recognition modules.

Change-Id: I317fa6d9839d876d45033015faef41e6291fde16
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: add face recognition framework
Inki Dae [Fri, 11 Mar 2022 05:18:05 +0000 (14:18 +0900)]
mv_machine_learning: add face recognition framework

[Version] 0.17.0-0
[Issue type] new feature

Added face recognition framework based the training engine, NNTrainer[1]

[CAPI Description]
mv_face_recognition_create
- Create face recognition handle.

mv_face_recognition_destroy
- Destroy a given handle.

mv_face_recognition_prepare
- Configure face recognition with a given configuration file,
  media-vision-config.json, and prepare some resources for
  face recognition service.

mv_face_recognition_register
- Register a given new face, which is an process to train the given face
  and generate a internal model.

mv_face_recognition_unregister
- Remove a given label relevant all resource data from feature vector,
  label files, and regenerate a internal model.

mv_face_recognition_inference
- Inference with a given face image and the trained model, and return
  proper label related to the face image.

[How to use CAPI]
Refer to FaceRecognitionTest test case of
test/testsuites/machine_learning/face_recognition/test_face_recognition.cpp

[1] https://github.com/nnstreamer/nntrainer

Change-Id: I8410b515ae90772185772b4b1c49ba0da7885f98
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoFix bugs
Tae-Young Chung [Tue, 22 Feb 2022 06:30:50 +0000 (15:30 +0900)]
Fix bugs

[Version] 0.16.2-0
[Issue type] bug fix

1. Fix INVARIANT_RESULT.OP_ZERO
 - In case of GetLandmarkType(), its' return values are limited to the
landmark type enumeration so that checking the return value of
GetLandmarkType() is meaningless
2. Check null after malloc()
3. Initialize member variable in class ctor

Change-Id: I0e23298ffcb7760eb86e24e37347eae5e992e421
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agomv_barcode: Fix log level
Kwang Son [Mon, 28 Feb 2022 00:26:59 +0000 (19:26 -0500)]
mv_barcode: Fix log level

return error code log should be error

[Version] 0.16.1-0
[Issue type] bug fix

Change-Id: I6c19709c56a6882807f8bfa1d0b3586d86e0010f
Signed-off-by: Kwang Son <k.son@samsung.com>
2 years agomv_machine_learning: refactoring to directory structure
Inki Dae [Thu, 24 Feb 2022 09:23:26 +0000 (18:23 +0900)]
mv_machine_learning: refactoring to directory structure

[Version] 0.16.0-0
[Issue type] code refactoring

Moved inference directory to top one and dropped mv_inference.

Old directory structure
------------------------
mv_machine_learning --- mv_inference --- inference

New directory structure
-----------------------
mv_machine_learnizng --- inference

Change-Id: I2d53aee269c95972e01086697245cb23aa8cf7b2
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomachine_learning: update YOLO-V5 inference
Tae-Young Chung [Wed, 16 Feb 2022 07:01:05 +0000 (16:01 +0900)]
machine_learning: update YOLO-V5 inference

[Version] 1.15.1-0
[Issue type] Update

Yolo-V5's output is 1x10647x85 in case of an input has [416, 416]. But, a converted dlc (snpe) from onnx has 1x13x13x255 without processing anchor box.
This commit includes a patch handling that kinds of models even though results are weird.

Change-Id: I5d53f72bc3896b92a5a59442190e2767887fd22d
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agocode refactoring to testsuites for machine learning
Inki Dae [Tue, 8 Feb 2022 10:24:50 +0000 (19:24 +0900)]
code refactoring to testsuites for machine learning

[Version] : 0.15.0-0
[Issue type] : new feature

Did code refactoring to test suites for machine learning by
dropping code duplication and providing user given target device type.

Change-Id: I7e9b8df902af9c45d6d878b28a4e77e127e16982
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: apply inference engine profiler support
Inki Dae [Fri, 28 Jan 2022 04:04:58 +0000 (13:04 +0900)]
mv_machine_learning: apply inference engine profiler support

[Version] : 0.14.0-0
[Issue type] : new feature

Applied inference engine profiler support which provides
a build option for disabling or enabling inference engine profiler
in runtime. In default the profiler is disabled.

Change-Id: Ie85c078c52bf6f23017e50d91158af9b4a37d113
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoface: fix typo
Tae-Young Chung [Fri, 28 Jan 2022 01:02:56 +0000 (10:02 +0900)]
face: fix typo

[Version] 0.13.2-0
[Issue type] bug fix

Change-Id: Ib24296ab0f428d001f718a6ba16c248e96b40982
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agotestsuite: Move helper functions before caller functions
Tae-Young Chung [Wed, 26 Jan 2022 01:28:37 +0000 (10:28 +0900)]
testsuite: Move helper functions before caller functions

[Version] 0.13.1-0
[Issue type] bug fix

Some functions call helper functions which are just defined behind them
without any declaration so that 'implicit declaration function' warning occurs.
Instead of declaring prototype, just move them to before the some function.

Change-Id: I5defa81ba22f39981eb97553dda78d3191d49ff0
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agomv_machine_learning: support YOLO-V5 inference
Tae-Young Chung [Tue, 25 Jan 2022 07:18:19 +0000 (16:18 +0900)]
mv_machine_learning: support YOLO-V5 inference

[Version] 0.13.0-0
[Issue type] new feature

Change-Id: I6dd5d13904e7c201693441494f87deba42f35924
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agomv_machine_learning: fix typo
Hyunil [Wed, 5 Jan 2022 08:45:05 +0000 (17:45 +0900)]
mv_machine_learning: fix typo

[Version] 0.12.1-0
[Issue type] fix typo

Change-Id: If058f551b3f5cdf92855d7e9ec95b1066d120c28
Signed-off-by: Hyunil <hyunil46.park@samsung.com>
2 years agodocs: fix typo 15/278315/1 accepted/tizen/unified/20220720.064101 submit/tizen/20220719.105718
Hyunsoo Park [Tue, 19 Jul 2022 04:37:16 +0000 (13:37 +0900)]
docs: fix typo

[Version] 0.12.13-0
[Issue type] doc update

Change-Id: Ib637bb2086da549499558e2fd8dafd032015438b
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years ago docs: fix typo 27/277827/1 accepted/tizen/unified/20220715.141311 submit/tizen/20220714.054227
Hyunsoo Park [Thu, 14 Jul 2022 02:20:08 +0000 (11:20 +0900)]
 docs: fix typo

 [Version] 0.12.12-0
 [Issue type] doc update

Change-Id: I660abc56ae1ae60ac39139197d1170fb8b9e88be
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years ago[ACR-1706] mv_image: add return value 65/277265/1
Hyunsoo Park [Mon, 4 Jul 2022 07:44:26 +0000 (16:44 +0900)]
[ACR-1706] mv_image: add return value

From this patch,
'mv_image_tracking_model_set_target' api could return MEDIA_VISION_ERROR_INVALID_OPERATION
when 'CascadeTracker' is null.

Change-Id: I73b50d9a5b9dd5a592370d1d7c30973402b661a0
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years agomv_image: Fix bugs 41/277241/1 submit/tizen/20220704.055720
Hyunsoo Park [Mon, 4 Jul 2022 04:59:13 +0000 (13:59 +0900)]
mv_image: Fix bugs

[Version] 0.12.11-0
[Issue type] bug fix

Return if 'CascadeTracker' is null.

Change-Id: Ica09a490d8dc9cd9268172d3c1a1a05d6cf1472a
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years agoFix implicit declaration function 13/275813/1
Kwanghoon Son [Thu, 2 Jun 2022 06:07:25 +0000 (02:07 -0400)]
Fix implicit declaration function

Use before function declaration and this makes wrong function usage.

[Version] : 0.12.10-0
[Issue type] : Bug fix

Change-Id: I2f8cc1386acd6c2d6afb098729140fe070725424
Signed-off-by: Kwanghoon Son <k.son@samsung.com>
2 years agoExclude line coverage test files 10/275810/1
Kwanghoon Son [Thu, 2 Jun 2022 05:43:32 +0000 (01:43 -0400)]
Exclude line coverage test files

[Version] : 0.12.9-0
[Issue type] : Bug fix

Change-Id: Ie3ed102e0bd30332cd4d5efaadc56425a8e84edd
Signed-off-by: Kwanghoon Son <k.son@samsung.com>
2 years agoChange gcov rpm packaging path 90/275490/3 accepted/tizen/unified/20220528.144400 submit/tizen/20220526.043714
Kwang Son [Tue, 24 May 2022 23:54:05 +0000 (19:54 -0400)]
Change gcov rpm packaging path

[Version] 0.12.8
[Issue type] Gcov

Change gcov package to show relative code path
gcov/obj/%{name} -> gcov/obj/%{name}/%{builddir}

Change-Id: I27895951a5e39d6b637ec9734a01b54f8679dfdc
Signed-off-by: Kwang Son <k.son@samsung.com>
2 years agodocs: Fix typo 50/275350/4 accepted/tizen/unified/20220525.210651 submit/tizen/20220523.060800
Hyunsoo Park [Mon, 23 May 2022 05:36:24 +0000 (14:36 +0900)]
docs: Fix typo

[Version] 0.12.7-0
[Issue type] Fix typo

Change-Id: I894675ef5904bd9d10a9b4d40218eafed81db989
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
2 years agoclean up manual file copy 65/274765/1
sangho park [Mon, 9 May 2022 05:53:43 +0000 (14:53 +0900)]
clean up manual file copy

[Version] 0.12.6-0
[Issue type] cleanup

delete hard copy commands in spec file and implement in CMakeLists.txt

Change-Id: I75841678480972f2a270b84db6fc533e006e1d5d
Signed-off-by: sangho park <sangho.g.park@samsung.com>
2 years agofix cmake_minimum_required() deprecation warning. 71/272971/2
sangho park [Tue, 29 Mar 2022 02:07:41 +0000 (11:07 +0900)]
fix cmake_minimum_required() deprecation warning.

[Version] : 0.12.5-0
[Issue type] : Bug fix

Change-Id: I6b178d31b204ebdc38860bd82ed942867782ee0f
Signed-off-by: sangho park <sangho.g.park@samsung.com>
2 years agoFix bugs 97/271497/4 accepted/tizen/unified/20220224.133705 submit/tizen/20220224.031004
Tae-Young Chung [Tue, 22 Feb 2022 06:30:50 +0000 (15:30 +0900)]
Fix bugs

[Version] 0.12.4-0
[Issue type] bug fix

1. Fix INVARIANT_RESULT.OP_ZERO
 - In case of GetLandmarkType(), its' return values are limited to the
landmark type enumeration so that checking the return value of
GetLandmarkType() is meaningless
2. Check null after malloc()
3. Initialize member variable in class ctor

Change-Id: I0e23298ffcb7760eb86e24e37347eae5e992e421
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agoface/image: Add opencv's tracking header files 60/269860/1 accepted/tizen/unified/20220125.121712 submit/tizen/20220124.072348
Tae-Young Chung [Fri, 21 Jan 2022 00:28:32 +0000 (09:28 +0900)]
face/image: Add opencv's tracking header files

[Version] 0.12.3-0
[Issue type] Update

From OpenCV-4.5.3, a tracking module is moved to contrib
and some tracking modes is changed to legacy.

Change-Id: Ibdfcc89abf5f10a8d874da81ab3fc4601cec5423
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agopackaging: Change gcov object install path 90/269690/1 accepted/tizen/unified/20220124.020029 submit/tizen/20220120.235212
Kwang Son [Thu, 20 Jan 2022 04:15:07 +0000 (23:15 -0500)]
packaging: Change gcov object install path

[Version] : 0.12.2-0
[Issue type] : feature request

Change gcov install path with append package prefix for better
directory structure.

Change-Id: Icf586a264e1f8a37db31c134e5e2ea9f82968e21
Signed-off-by: Kwang Son <k.son@samsung.com>
2 years agotest/machine_learning: fix device attribute type 19/269219/2
Inki Dae [Tue, 11 Jan 2022 11:26:14 +0000 (20:26 +0900)]
test/machine_learning: fix device attribute type

[Version] : 0.12.1-0
[Issue type] : bug fix

Fixed device attribute type. Legacy device attribute type has been
deprecated so use new one.

Change-Id: Ia31a5be01595e6d39f3dff783e5b4d0982987e86
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agoinference: support movenet 94/268794/2 accepted/tizen/unified/20220110.140009 submit/tizen/20220105.080154 submit/tizen/20220105.081745
Tae-Young Chung [Mon, 20 Dec 2021 09:01:24 +0000 (18:01 +0900)]
inference: support movenet

[Version] 0.12.0
[Issue type] new feature

Movenet models with variants, Lightning3/4 and Thunder3/4, are supported.
Models can be downloaded from tfhub.dev and
meta files for the models are added as examples.
In addition, testsuite with those Movenet models are updated.

Change-Id: I1c526cba688842e408fad1c84ba7dff4d0320e80
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
2 years agomv_maching_learning: deprecate MLAPI and MAX types 41/268741/9
Inki Dae [Fri, 31 Dec 2021 01:50:53 +0000 (10:50 +0900)]
mv_maching_learning: deprecate MLAPI and MAX types

[Version] : 0.11.0-0
[Issue type] : cleanup

Deprecated MV_INFERENCE_BACKEND_MLAPI and MV_INFERENCE_BACKEND_MAX types.
MV_INFERENCE_BACKEND_MLAPI is not backend type but other API framework,
and another may break the binary compatibility when adding a new enumeration
value because MAX value is different from one of the binary
So drop these two types.

Change-Id: I391cd0d4b713e3d35fe263f1567f4cea3df60630
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agotest/machine_learning: add snpe model support 41/268641/1
Inki Dae [Tue, 28 Dec 2021 10:31:33 +0000 (19:31 +0900)]
test/machine_learning: add snpe model support

[Version] : 0.10.0-0
[Issue type] : new feature

Added a test case for SNPE engnie with dlc model.

Change-Id: I7d20f9974300130ddeaf4e8eb77482d89dee0b9d
Signed-off-by: Inki Dae <inki.dae@samsung.com>
2 years agomv_machine_learning: add SNPE inference engine support 32/267032/2
Inki Dae [Wed, 24 Nov 2021 08:39:23 +0000 (17:39 +0900)]
mv_machine_learning: add SNPE inference engine support

[Version] : 0.9.0-0
[Issue type] : new feature

Change-Id: Iabfc4932d808296c8941bc8bbacf88b9fbe09616
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to property parsing 17/265417/3
Inki Dae [Mon, 18 Oct 2021 07:06:27 +0000 (16:06 +0900)]
mv_machine_learning: code refactoring to property parsing

Did code refactoring to propery parsing of BoxInfo and Landmark classes.

What this patch did,
 - Move parseLabel, ParseNumber, ParseDecodeInfo functions from
   OutputMetadata class to BoxInfo one because these properties
   are related to BoxInfo class not OutputMetadata class.
 - Move parseDisplayment, parseEdgeMap, parseDecodeInfo functions from
   OutputMetadata class to Landmark one because these properties
   are related to Landmark class not OutputMetadata class.

This is just one step for next code refactoring.

Change-Id: I0a17ee29f492bee53acacab339ba746a44495790
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to OutputMetadata module 41/265341/3
Inki Dae [Thu, 14 Oct 2021 06:43:59 +0000 (15:43 +0900)]
mv_machine_learning: code refactoring to OutputMetadata module

OutputMetadata.h and cpp files contain many classes and relevant
code so it makes maintenance too hard.

The biggest change of this refactoring is to separate many classes
in OutputMetadata file into each one. And also it changes unnecessary
class type to struct including several cleanups such as code sliding,
renaming and dropping unnessary code.

Change-Id: I0ce677d333ce3a3e7212f7d26a20b6cf77bc7a9a
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code cleanup to Inference class 87/265287/1
Inki Dae [Wed, 13 Oct 2021 09:55:41 +0000 (18:55 +0900)]
mv_machine_learning: code cleanup to Inference class

Cleaned up Inference class.

What this patch does,
 - code sliding.
 - change variable name to meaningful one such as 'n' to 'output_idx'.
 - drop unnecessary variable.
 - use same parameter name for Get*Results functions.

Change-Id: I47ac3eb241116174e4a6a7bc2a1b90ab9378de25
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to input and output metadata classes 73/265073/6
Inki Dae [Thu, 7 Oct 2021 10:03:23 +0000 (19:03 +0900)]
mv_machine_learning: code refactoring to input and output metadata classes

Did code refactoring to InputMetadata and OuputMetadata.

The biggest change of this patch is to delegate all requests
of various metadata relevant classes to InputMetadata and
OutputMetadata classes to mitigate Inference class dependency of
Metadata things.

In addition, it makes several classes declared in BoxInfo class to get out
, and changes class to struct in case that class functionality isn't needed.

And some cleanup and code sliding.

This is just one step for next code refactoring.

Change-Id: I92c4e2b2c77499246c56a237282c05262550301a
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code cleanup to GetFacialLandMarkDetectionResults function 72/265072/3
Inki Dae [Thu, 7 Oct 2021 06:30:36 +0000 (15:30 +0900)]
mv_machine_learning: code cleanup to GetFacialLandMarkDetectionResults function

Did code cleanup to GetFacialLandMarkDetectionResults function
by doing,
   - code sliding
   - change variable name to meaningful name for readability.

Change-Id: Ib59786c085c8202a1f7d9eb85a01d528220c728f
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agotest/machine_learning/inference: add test cases for legacy path 59/264859/2
Inki Dae [Thu, 30 Sep 2021 10:45:34 +0000 (19:45 +0900)]
test/machine_learning/inference: add test cases for legacy path

Added test cases for legacy path of inference engine, which uses
user-given model information instead of ones from json file.

As for this, this patch has a little bit code refactoring which
uses parameterized test instead of fixture one of google test
to decide API path in runtime - json or legacy.

This patch enlarges exsiting test coverage from 119 to 132 test cases.

[==========] 132 tests from 6 test suites ran. (49021 ms total)
[  PASSED  ] 132 tests.

Change-Id: I9829725aad8037cbe5a82d50e7790a3e7a6bfe6b
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to GetClassificationResults function 18/264818/1
Inki Dae [Wed, 29 Sep 2021 09:37:43 +0000 (18:37 +0900)]
mv_machine_learning: code refactoring to GetClassificationResults function

Did code refactoring to GetClassificationResults member function of Inference class
for next code refactoring.

What this patch did are,
 - drop code dependency of mPreProc from Inference class.
 - drop code duplication of getting image classification result by
   making same code to be used.
 - and code sliding.

Change-Id: I721dc3081f730fcdb75dbe91ad743f906e076ea4
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to GetPoseLandmarkDetectionResults function 29/264729/5
Inki Dae [Mon, 27 Sep 2021 09:27:31 +0000 (18:27 +0900)]
mv_machine_learning: code refactoring to GetPoseLandmarkDetectionResults function

Did code refactoring to GetPoseLandmarkDetectionResults function of
Inference class by dropping Inference class dependency to mPoseResult object.

What this path did are,
 - use smart pointer instead of new operator, and make postResult
   object to be valid only in mv_inference_pose_landmark_detect_open
   function context.
 - drop some unnecessary local variables.
 - and code sliding.

The biggest change of this patch is to drop Inference class dependency
of mPoseResult object for next code refactoring. As as result,
it drops one brace from GetPoseLandmarkDetectionResults function and makes
the function to be more simplified for code readability enhancement.

Change-Id: I25d1a175571d407404c088713a14e485496b2920
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning:code refactoring to Inference.cpp 34/264734/3
Inki Dae [Mon, 27 Sep 2021 07:49:21 +0000 (16:49 +0900)]
mv_machine_learning:code refactoring to Inference.cpp

Did code refactoring to Inference.cpp module for next code
refactoring work.

What this patch did are,
 - just drop all unnecessary members  of Inference class
 - and gather some spreaded code to relevant place.

Change-Id: I2fa00af5b0b8701c4bdbbf62d07a2f9f2ef2bc85
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agotest: Add meta files 29/264529/1 sandbox/dfs submit/tizen/20210924.044804
Tae-Young Chung [Fri, 24 Sep 2021 03:03:11 +0000 (12:03 +0900)]
test: Add meta files

Change-Id: I765ba21d3be44ac6305c6c5ffcdddf4ccdac746b
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agomv_machine_learning: make IsXXX functions simplified 09/264409/2
Inki Dae [Fri, 17 Sep 2021 06:37:21 +0000 (15:37 +0900)]
mv_machine_learning: make IsXXX functions simplified

This patch changes existing code to just "return condition",
and declare IsJsonFile function as static function because
this function is used only in this module.

Change-Id: I86d8959ef32e382e4a0be76cbdc7ff781cede6e4
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: code refactoring to checking custom device type 00/264400/5
Inki Dae [Fri, 17 Sep 2021 04:40:52 +0000 (13:40 +0900)]
mv_machine_learning: code refactoring to checking custom device type

This patch does code refactoring for enhancing code readiability
of the condition which checks if config type path is required or
not according to given backend and target device types.

As for this, this patch adds IsValidBackendType fucntion to check if
a given backend type is valid or not, IsConfigFilePathRequired
function to check if config file path is required or not, and
makes these functions to be used instead of existing complicated condition.

Change-Id: I195e3caeb205b159e2128a14c2b4b6516e7b952d
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agotest: Add TestPoseLandmarkDetection 90/264390/1
Kwang Son [Fri, 17 Sep 2021 01:04:05 +0000 (21:04 -0400)]
test: Add TestPoseLandmarkDetection

Change-Id: Iebbfda014bbf9e6bf0db68c79b675f9879b7de5d
Signed-off-by: Kwang Son <k.son@samsung.com>
3 years agotest: Add TestFaceLandmarkDetection 40/264240/2
Kwang Son [Wed, 15 Sep 2021 01:29:16 +0000 (21:29 -0400)]
test: Add TestFaceLandmarkDetection

Change-Id: Ia54e4183c0a50044441b96fb991abe19741f8b3a
Signed-off-by: Kwang Son <k.son@samsung.com>
3 years agoMerge "mv_machine_learning: code cleanup for refactoring" into tizen_devel
Inki Dae [Wed, 15 Sep 2021 07:07:51 +0000 (07:07 +0000)]
Merge "mv_machine_learning: code cleanup for refactoring" into tizen_devel

3 years agometa-template: fix json parse error 47/264247/1
Kwang Son [Wed, 15 Sep 2021 04:15:07 +0000 (00:15 -0400)]
meta-template: fix json parse error

Change-Id: I03607e65fa52f6f048333943a97cb5cd23576ee1
Signed-off-by: Kwang Son <k.son@samsung.com>
3 years agomv_machine_learning: code cleanup for refactoring 95/264195/2
Inki Dae [Tue, 14 Sep 2021 06:55:36 +0000 (15:55 +0900)]
mv_machine_learning: code cleanup for refactoring

Cleaned up mv_inference_open.cpp module for code refactoring.
What cleanups are,
 - correct variable name which doesn't clarify what the name means.
 - change goto label name to be meaningful with resource release issue.
 - drop unnecessary code.
 - move code so that relevant code could be grouped together.

This patch is just one step of what we have to for code refactoring.

Change-Id: Ia744508e0318d6558a0f4901a90d7ac8a14634e2
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agotest: set meta file path accepted/tizen/6.5/unified/20211028.094500 accepted/tizen/unified/20210909.123030 submit/tizen/20210909.010804 submit/tizen_6.5/20211028.161801 tizen_6.5.m2_release
Inki Dae [Tue, 7 Sep 2021 07:21:52 +0000 (16:21 +0900)]
test: set meta file path

Since commit-id c4e918cda1ef97fdd5e5cf74dd056b65407ab671,
Mediavision needs a json file per a model.

This patch makes json file path using a given model file path
and sets it in case of reference model.

Change-Id: I51c9721fbebeb6c79abeb67faa193b9580bbae95
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: drop GetResults function
Inki Dae [Tue, 7 Sep 2021 05:12:24 +0000 (14:12 +0900)]
mv_machine_learning: drop GetResults function

This function isn't used so drop it.

Change-Id: I4c3b95bb3716d14bd42dc1531b9be6f33d958a62
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: introduce allocate function for TensorBuffer class
Inki Dae [Tue, 7 Sep 2021 03:59:07 +0000 (12:59 +0900)]
mv_machine_learning: introduce allocate function for TensorBuffer class

Added allocate function to allocate buffers for input and output tensors.
This patch drops code duplication - same code was used to allocate
input and output tensor buffers.

Change-Id: I0cfb46633a9a3fe74c734307ac9c15b5df064e63
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agotest: set meta file path
Inki Dae [Tue, 7 Sep 2021 07:21:52 +0000 (16:21 +0900)]
test: set meta file path

Since commit-id c4e918cda1ef97fdd5e5cf74dd056b65407ab671,
Mediavision needs a json file per a model.

This patch makes json file path using a given model file path
and sets it in case of reference model.

Change-Id: I51c9721fbebeb6c79abeb67faa193b9580bbae95
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: drop GetResults function
Inki Dae [Tue, 7 Sep 2021 05:12:24 +0000 (14:12 +0900)]
mv_machine_learning: drop GetResults function

This function isn't used so drop it.

Change-Id: I4c3b95bb3716d14bd42dc1531b9be6f33d958a62
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agomv_machine_learning: introduce allocate function for TensorBuffer class
Inki Dae [Tue, 7 Sep 2021 03:59:07 +0000 (12:59 +0900)]
mv_machine_learning: introduce allocate function for TensorBuffer class

Added allocate function to allocate buffers for input and output tensors.
This patch drops code duplication - same code was used to allocate
input and output tensor buffers.

Change-Id: I0cfb46633a9a3fe74c734307ac9c15b5df064e63
Signed-off-by: Inki Dae <inki.dae@samsung.com>
3 years agoMerge "Skip ParseMetadata() if a metadata json file doesn't exist" into tizen accepted/tizen/unified/20210909.002355 submit/tizen/20210908.100700
Tae-Young Chung [Wed, 8 Sep 2021 10:06:17 +0000 (10:06 +0000)]
Merge "Skip ParseMetadata() if a metadata json file doesn't exist" into tizen

3 years agoFix coverity issue(RESOURCE_LEAK/UNINIT) 49/263649/3 submit/tizen/20210908.064001
Hyunsoo Park [Tue, 7 Sep 2021 10:44:13 +0000 (19:44 +0900)]
Fix coverity issue(RESOURCE_LEAK/UNINIT)

Change-Id: I13e283e496dc815cf5a168df3a365b83f4c3cf83
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
3 years agoSkip ParseMetadata() if a metadata json file doesn't exist 67/263667/1
Tae-Young Chung [Wed, 8 Sep 2021 01:14:23 +0000 (10:14 +0900)]
Skip ParseMetadata() if a metadata json file doesn't exist

Change-Id: Ibfc01bd79fe300447034f7504c4573abec47223a
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoAdd MV_INFERENCE_MODEL_META_FILE_PATH attribute and deprecate some attributes which... 65/263465/4 submit/tizen/20210907.045211
Tae-Young Chung [Fri, 3 Sep 2021 03:58:27 +0000 (12:58 +0900)]
Add MV_INFERENCE_MODEL_META_FILE_PATH attribute and deprecate some attributes which can be described in a metafile

Deprecate attributes
- MV_INFERENCE_MODEL_MEAN_VALUE
- MV_INFERENCE_MODEL_STD_VALUE
- MV_INFERENCE_INPUT_TENSOR_WIDTH
- MV_INFERENCE_INPUT_TENSOR_HEIGHT
- MV_INFERENCE_INPUT_TENSOR_CHANNELS
- MV_INFERENCE_INPUT_DATA_TYPE
- MV_INFERENCE_INPUT_NODE_NAME
- MV_INFERENCE_INPUT_OUTPUT_NODE_NAMES
- MV_INFERENCE_OUTPUT_MAX_NUMBER
- MV_INFERENCE_CONFIDENCE_THRESHOLD

Change-Id: I340b6c2fa3872dd682bc2f60d45007ffb87a46b5
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoRemove unnecessary dim_type metadata 55/263455/1
Tae-Young Chung [Fri, 3 Sep 2021 01:14:32 +0000 (10:14 +0900)]
Remove unnecessary dim_type metadata

Change-Id: Ied3b116ad0e34becd9e87caafa59de8069f2639f
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange offset's shape_type and displacement's shape_type and type to string 35/263435/1
Tae-Young Chung [Thu, 2 Sep 2021 07:04:05 +0000 (16:04 +0900)]
Change offset's shape_type and displacement's shape_type and type to string

Change-Id: Ide16c57e44532d10e9633a1399c6d787991baf33
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange Landmark's landmark_type, landmark_coordinate, and decoding_type to string 05/263305/2
Tae-Young Chung [Tue, 31 Aug 2021 08:09:49 +0000 (17:09 +0900)]
Change Landmark's landmark_type, landmark_coordinate, and decoding_type to string

A user can use the string while understanding purpose of
landmark_type, landmark_coordinate, and decoding_type metadata.
The string values are parsed and converted to enumeration type
inference_landmark_type_e, inference_landmark_coordinate_type_e,
and inference_landmark_decoding_type_type_e, respectively.

Change-Id: Ia3a4098213a712ffd1927838d7b9931d1edfcbd0
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoFix missing goto _ERROR_ 86/263286/1
Tae-Young Chung [Tue, 31 Aug 2021 00:08:29 +0000 (09:08 +0900)]
Fix missing goto _ERROR_

Change-Id: I07f7cc7dfd90677b6d8fdbe3f82068490406562b
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange int CheckSupportedTargetDevice() to bool IsTargetDeviceSupported() and print... 79/263279/2
Tae-Young Chung [Mon, 30 Aug 2021 12:25:11 +0000 (21:25 +0900)]
Change int CheckSupportedTargetDevice() to bool IsTargetDeviceSupported() and print format of targetDevices.

Change-Id: I1f9e92634a09f1f93222762cea93f0d25a71aeee
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange behavior before and after pInfer->Bind() 47/263047/1
Tae-Young Chung [Wed, 25 Aug 2021 04:57:13 +0000 (13:57 +0900)]
Change behavior before and after pInfer->Bind()

In Inference::Bind(), pInfer->Bind(), mConfig.mTargetTypes is used.
But, mConfig.mTargetTypes is set by pInfer->ConfigureTargetTypes and
pInfer->ConfigureTargetDevices after pInfer->Bind().
So we change the behavior as follows.
before pInfer->Bind(),
mConfig.mTargetTypes is set by pInfer->ConfigureTargetTypes().
After pInfer->Bind(),
pInfer->CheckSupportedTargetDevice() just checks the mConfig.mTargetTypes
is supported or not.

Change-Id: I79d2dc6372aac0c503783fb9fef2b4eccecbb018
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoFix '!=' operator to '&' for supported_accel_devices bitwise comparison 90/262990/1
Tae-Young Chung [Tue, 24 Aug 2021 08:46:39 +0000 (17:46 +0900)]
Fix '!=' operator to '&' for supported_accel_devices bitwise comparison

Change-Id: Ief7d7deebcc9e2fcac83ff98f9b01f0c40e09f38
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoFix mv_engine_config_set_int_attribute() to mv_engine_config_set_string_attribute() 74/262974/1
Tae-Young Chung [Tue, 24 Aug 2021 07:04:37 +0000 (16:04 +0900)]
Fix mv_engine_config_set_int_attribute() to mv_engine_config_set_string_attribute()

Change-Id: Ic6e10c1739746ca01eb29da68901dfd4dc0762db
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoAdd return value handling 52/262952/1
Tae-Young Chung [Tue, 24 Aug 2021 02:29:59 +0000 (11:29 +0900)]
Add return value handling

error should be returned when exception is catched and
is handled properly.

Change-Id: If669515f2b573d3c96165647bd8eaca852d373b1
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange BoxInfo's box_coordinate, decoding_type and NMSParam's mode to string 89/262889/2
Tae-Young Chung [Mon, 23 Aug 2021 06:38:13 +0000 (15:38 +0900)]
Change BoxInfo's box_coordinate, decoding_type and NMSParam's mode to string

A user can use the string while understanding purpose of
box_coordinate, decoding_type, and mode metadata.
The string values are parsed and converted to enumeration type
inference_box_coordinate_type_e, inference_box_decoding_type_e,
and inference_box_nms_type_e, respectively.

Change-Id: I4419a288f070456c9f8d411feed7706cd0f806f0
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoFix bug in OjbectDecdoder::decode() 15/262815/2
Tae-Young Chung [Fri, 20 Aug 2021 05:57:30 +0000 (14:57 +0900)]
Fix bug in OjbectDecdoder::decode()

GetType should be GetDecodingType().
decoding is based on decoding_type, not box_type.

Change-Id: I4ed7cc64ed0f9a1503279fca7d20c6e0f8414ed7
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange BoxInfo's box_type in outputmetadata to string 14/262814/4
Tae-Young Chung [Fri, 20 Aug 2021 05:43:35 +0000 (14:43 +0900)]
Change BoxInfo's box_type in outputmetadata to string

box_type in BoxInfo outputmetadata was integer value.
It is changed to string, ORIGIN_LEFTTOP and ORIGIN_CENTER.
A user can use the string while understading purpose of box_type metadata.
The string values are parsed and converted to enumeration
INFERENCE_BOX_TYPE_ORIGIN_LEFTTOP or INFERENCE_BOX_TYPE_ORIGIN_CENTER.

Change-Id: Icab71b26ca3c96652c23a584eb8a3ec5c6d7d44b
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agotest: Add FaceDetection test 85/262785/1
Kwang Son [Thu, 19 Aug 2021 23:02:08 +0000 (19:02 -0400)]
test: Add FaceDetection test

Change-Id: Ice45ea84dcdb2c2fd1348e1c0c37210d3c94b30c
Signed-off-by: Kwang Son <k.son@samsung.com>
3 years agoMerge "test: break condition" into tizen_devel accepted/tizen/unified/20210820.101812 submit/tizen/20210820.005508
kwang son [Thu, 19 Aug 2021 22:46:51 +0000 (22:46 +0000)]
Merge "test: break condition" into tizen_devel

3 years agoChange pkgconfig(gtest) to gtest-devel 73/262773/2
Tae-Young Chung [Thu, 19 Aug 2021 10:02:35 +0000 (19:02 +0900)]
Change pkgconfig(gtest) to gtest-devel

gtest which is used in GBM is older version without a pc file.
To prevent build failure, use gtest-devel instead of pkgconfig(gtest)
in spec file.

Change-Id: I67f7e5f02288559bdf886dca33f47f68be8d7552
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agoChange ScoreInfo's score_type in outputmetadata to string 42/262742/4
Tae-Young Chung [Thu, 19 Aug 2021 03:26:58 +0000 (12:26 +0900)]
Change ScoreInfo's score_type in outputmetadata to string

score_type in ScoreInfo outputmetadata was integer value which an user doesn't understand
even though it was provided for the user convenience.
So they are changed to string and user can understand what they mean.
The string values are parsed and converted to proper enumerations.

Change-Id: I9f09bb6f2c534ff5a8f5084a3e2413514b075f5d
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>
3 years agotest: break condition 47/262747/1
Kwang Son [Thu, 19 Aug 2021 05:59:07 +0000 (01:59 -0400)]
test: break condition

Change-Id: I9994f71ad8627dd4a3d6792e87b9602651ba76b4
Signed-off-by: Kwang Son <k.son@samsung.com>
3 years agoMerge "test: Add ObjectDetection test" into tizen_devel
kwang son [Thu, 19 Aug 2021 05:46:58 +0000 (05:46 +0000)]
Merge "test: Add ObjectDetection test" into tizen_devel

3 years agoAdd Quantize() to PreProcess 01/262701/2
Tae-Young Chung [Wed, 18 Aug 2021 06:36:07 +0000 (15:36 +0900)]
Add Quantize() to PreProcess

Change-Id: I4df22f0ec4e1bcf663459083ce77f70d91c7a600
Signed-off-by: Tae-Young Chung <ty83.chung@samsung.com>