OpenGVLab / OpenGVLab/InternVideo
Bug in ViCLIP's implementation + demo
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2.4k
- Forks
- 160
- PR merge metrics
- No merged PRs in 30d
Description
In https://github.com/OpenGVLab/InternVideo/blob/6264cc85f72e38dce7e38549182d0369c50cde00/Data/InternVid/viclip/viclip.py#L140-L143 a single video is converted to a batch of 8 images (instead of a batch of one video). This bug is transferred to the demo in https://github.com/OpenGVLab/InternVideo/blob/6264cc85f72e38dce7e38549182d0369c50cde00/Data/InternVid/viclip/__init__.py#L71 where the outputs for the incorrect batch size of 8 are thrown out and only the first frame's outputs are used.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with Data/InternVid/viclip/viclip.py lines 140-143 and inspect how one video is converted into a batch; then follow the corresponding use in Data/InternVid/viclip/init.py line 71. Verify the expected batch and output shapes, and confirm the demo uses the complete video's outputs rather than discarding all but the first frame.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- computer-vision
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Clearly specified
- Newbie friendliness
- 48/100