facebookresearch / facebookresearch/sam2
About _apply_non_overlapping_constraints
Open
- Dominant language
- Jupyter Notebook
- Stars
- 19.9k
- Forks
- 2.5k
- PR merge metrics
- No merged PRs in 30d
Description
I don't quite understand exactly what this function does. Why are these constraints on the prediction mask logits under the batch dimension when batch_size >1.
Hopefully someone can explain the following examples of what it is used for. Thank you very much.
Contributor guide
Assessment
This issue has not been assessed yet.