VideoSeg-R1:Reasoning Video Object Segmentation via Reinforcement Learning
Zishan Xu, Yifu Guo, Yuquan Lu, Fengyu Yang, Junxin Li, Lihua Cai
Abstract
Traditional video reasoning segmentation methods rely on supervised fine-tuning, which limits generalization to out-of-distribution scenarios and lacks explicit reasoning. To address this, we propose VideoSeg-R1, the first framework to introduce reinforcement learning into video reasoning segmentation. It adopts a decoupled architecture that formulates the task as joint referring image segmentation and video mask propagation. It comprises three stages: (1) A hierarchical text-guided frame sampler to emulate human attention; (2) A reasoning model that produces spatial cues along with explicit reasoning chains; and (3) A segmentation-propagation stage using SAM2 and XMem. A task difficulty-aware mechanism adaptively controls reasoning length for better efficiency and accuracy. Extensive evaluations on multiple benchmarks demonstrate that VideoSeg-R1 achieves state-of-the-art performance in complex video reasoning and segmentation tasks.
BibTeX
@inproceedings{aaai2026_videosegr1reason,
title = {VideoSeg-R1:Reasoning Video Object Segmentation via Reinforcement Learning},
author = {Zishan Xu and Yifu Guo and Yuquan Lu and Fengyu Yang and Junxin Li and Lihua Cai},
booktitle = {AAAI 2026},
year = {2026}
}