facebookresearch / facebookresearch/sam3
Tracking long videos
Open
- Dominant language
- Python
- Stars
- 11.7k
- Forks
- 1.8k
- PR merge metrics
- No merged PRs in 30d
Description
When tracking long videos, a large amount of data is cached, which can lead to very high video memory usage over time. Is there any good solution?
Contributor guide
Research direction
No file, test, or entry point is named. Start by locating the long-video tracking implementation and reproducing the reported video-memory growth, then trace which cached data accumulates over time. Done means the tracking workflow no longer shows unbounded video-memory usage on long videos, with a regression test or measured reproduction demonstrating the result.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- computer-vision, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100