search

cvlab-kaist / CAT-Seg

Official Implementation of "CAT-Seg🐱: Cost Aggregation for Open-Vocabulary Semantic Segmentation"
https://ku-cvlab.github.io/CAT-Seg/
MIT License
270 stars 25 forks source link

SAM + CAT-Seg details #15

Closed saikatdutta closed 7 months ago

saikatdutta commented 10 months ago

Hi, where can I find details of SAM + CAT-Seg? Thanks in advance!

hsshin98 commented 7 months ago

Hi, the details for SAM + CAT-Seg is not specified in our paper as we devised it as an independent demo for applications, but it has similar implementation to this repo: https://github.com/fudan-zvg/Semantic-Segment-Anything. We first predict open-vocabulary semantic segmentation results with CAT-Seg, and use this to classify the masks from SAM. This is done by considering the majority class in terms of number of pixels within the SAM mask as the predicted class, as in the aforementioned repository.