mv_machine_learning: add add async API for landmark detection task group
authorInki Dae <inki.dae@samsung.com>
Wed, 27 Sep 2023 06:52:36 +0000 (15:52 +0900)
committerKwanghoon Son <k.son@samsung.com>
Wed, 25 Oct 2023 01:54:03 +0000 (10:54 +0900)
commit7bc9d223fe3dec2f5faba4484b9adb9356b35179
tree3227ac7f86bb894ca437219dd1f79c32ce2ad8e3
parentcfe8c4f81415156e942242ad9edc54d4b3913e66
mv_machine_learning: add  add async API for landmark detection task group

[Issue type] : new feature

Add two asynchronous API, mv_facial_landmark_inference_async and
mv_pose_landmark_inference_async, for landmark detection task group
using AsyncManger class which is a common class for asynchronous API support
including test case.

In addition, this patch drops the use of mutex lock due to the dead lock issue
as known-issue mentioned already through the patch[1]

[1] commit-id : fdbd84f1ff3b30c9f18598b65da777f65611e4cb

Change-Id: Ifbf5a58d5053ca0226caff14913f4a0d3ba1c45e
Signed-off-by: Inki Dae <inki.dae@samsung.com>
12 files changed:
include/mv_facial_landmark_internal.h
include/mv_pose_landmark_internal.h
mv_machine_learning/landmark_detection/include/landmark_detection.h
mv_machine_learning/landmark_detection/include/landmark_detection_type.h
mv_machine_learning/landmark_detection/src/facial_landmark_adapter.cpp
mv_machine_learning/landmark_detection/src/landmark_detection.cpp
mv_machine_learning/landmark_detection/src/mv_facial_landmark.cpp
mv_machine_learning/landmark_detection/src/mv_pose_landmark.cpp
mv_machine_learning/landmark_detection/src/pose_landmark_adapter.cpp
packaging/capi-media-vision.spec
test/testsuites/machine_learning/landmark_detection/CMakeLists.txt
test/testsuites/machine_learning/landmark_detection/test_landmark_detection_async.cpp [new file with mode: 0644]