UX-Decoder / Semantic-SAM

[ECCV 2024] Official implementation of the paper "Semantic-SAM: Segment and Recognize Anything at Any Granularity"
2.33k stars 112 forks source link

Can you release an inference script? #27

Open shenkev opened 1 year ago

shenkev commented 1 year ago

Can you release a script that just takes 1 image or a folder of images and segments? With command line arguments for all the inference hyperparameters?

TengliEd commented 1 year ago

Glad to test our own data especially those without GT mask

FengLi-ust commented 1 year ago

You can try our auto_generation demo, which takes one image to segment everything and does not require GT masks.

FengLi-ust commented 1 year ago

Hi, we have released a simple getting started code for you to inference more conveniently.