facebookresearch / facebookresearch/sam2
sam2 stable inference
Open
- Dominant language
- Jupyter Notebook
- Stars
- 19.9k
- Forks
- 2.5k
- PR merge metrics
- No merged PRs in 30d
Description
I am using sam2 for realtime application and i need it to run in a stable inference speed.
I tried a lot of different methods but i get a lot of inference spikes without a specific pattern.
Is there anything i can do?
I tried efficient tracking and edgetam but i have the same problem.
Contributor guide
Assessment
This issue has not been assessed yet.