search

facebookresearch / segment-anything

The repository provides code for running inference with the SegmentAnything Model (SAM), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.
Apache License 2.0
46.75k stars 5.54k forks source link

Performance of CNN-based Segmentation Model using SA-1B #297

Closed Jayden9912 closed 1 year ago

Jayden9912 commented 1 year ago

Hi. May I know has anyone try to train any CNN-based segmentation model using the dataset SA-1B? I am curious what will be its performance in particular the zero-shot generalization.

ryouchinsa commented 1 year ago

The Fast Segment Anything Model(FastSAM) is a CNN Segment Anything Model trained using only 2% of the SA-1B dataset published by SAM authors.

Jayden9912 commented 1 year ago

That's cool. Thanks.