ReVSeg: Incentivizing the Reasoning Chain for Video Segmentation with RL

1 points| SweetSoftPillow | 2 months ago |arxiv.org