-
Notifications
You must be signed in to change notification settings - Fork 3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Allow to upload tracks from COCO formats #6969
Conversation
tests/python/rest_api/test_tasks.py
Outdated
task.import_annotations(format_name, dataset_file) | ||
|
||
imported_annotations = task.get_annotations() | ||
print(imported_annotations) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Need to remove
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Removed
Codecov Report
@@ Coverage Diff @@
## develop #6969 +/- ##
===========================================
+ Coverage 82.58% 82.59% +0.01%
===========================================
Files 360 360
Lines 38914 38928 +14
Branches 3570 3570
===========================================
+ Hits 32137 32154 +17
+ Misses 6777 6774 -3
|
This PR resolves the following problems: 1. During import we ignore track_id attribute for annotations if these annotations do not come from one of these formats `('cvat', 'datumaro', 'sly_pointcloud')` this prevents to import tracks in coco format -> this PR extends the list of "track formats" 2. When we convert a dataset from CVAT representation to Datumaro Dataset in all possible situations only the first shape of the skeleton track is a keyframe, all other shapes will be as marked as not keyframes it leads to problems when we try to import such tracks that were exported from CVAT -> this PR adds the following line of code: set `keyframe = True` for skeletons if at least one point of this skeleton is keyframe 3. During import skeleton tracks our code always expects to see a "keyframe" attribute for the skeleton's element, but not all formats support attributes for skeleton points (for example COCO Keypoints doesn't allow to have attributes for each skeleton's point). And if the server doesn't see this attribute for some element we ignore this element -> this PR adds the following logic for skeleton track: if skeleton's elements don't have a keyframe attribute at all then consider such skeleton elements as keyframes = True
Motivation and context
This PR resolves the following problems:
('cvat', 'datumaro', 'sly_pointcloud')
this prevents to import tracks in coco format -> this PR extends the list of "track formats"keyframe = True
for skeletons if at least one point of this skeleton is keyframeHow has this been tested?
Checklist
develop
branch(cvat-canvas,
cvat-core,
cvat-data and
cvat-ui)
License
Feel free to contact the maintainers if that's a concern.