AI RESEARCH
Learning Trajectory-Aware Multimodal Large Language Models for Video Reasoning Segmentation
arXiv CS.CV
•
ArXi:2603.21488v1 Announce Type: new The prosperity of Multimodal Large Language Models (MLLMs) has stimulated the demand for video reasoning segmentation, which aims to segment video objects based on human instructions. Previous studies rely on unidirectional and implicit text-trajectory alignment, which struggles with trajectory perception when faced with severe video dynamics. In this work, we propose TrajSeg, a simple and unified framework built upon MLLMs. Concretely, we