facebookresearch / facebookresearch/sam2
Output Difference Between SAM-base and SAM2.1_hiera_base_plus
Open
- Dominant language
- Jupyter Notebook
- Stars
- 19.9k
- Forks
- 2.5k
- PR merge metrics
- No merged PRs in 30d
Description
I noticed a significant difference in the output detail between SAM-base and SAM2.1_hiera_base_plus when using the same settings:
- `pred_iou_thresh: 0.75`
- `stability_score_thresh: 0.75`
SAM-base often produces more detailed output compared to SAM2.1_hiera_base_plus. Is this expected behavior, or should I investigate further? Thank you for your time and assistance!
This is mask generated by SAM-base:
And this mask generated by SAM2:
Contributor guide
Assessment
This issue has not been assessed yet.