Can one box prompt be used to segment all the objects of the same category in an image?

facebookresearch / sam2

The repository provides code for running inference with the Meta Segment Anything Model 2 (SAM 2), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.

Apache License 2.0

12.14k stars 1.1k forks source link

Can one box prompt be used to segment all the objects of the same category in an image? #408

Open lokvke opened 1 week ago

lokvke commented 1 week ago

For example, there are many people in an image, and I only use one of the people's box as prompt, can I segment all the people in the image?

catalys1 commented 4 days ago

SAM2 is designed for instance segmentation. I don't know of a way to use it for what you're describing, which is more akin to semantic segmentation.