VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer Authority

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Qiu, Chenhao, Zhang, Yechao, Luo, Xin, Song, Shien, Liu, Xusheng
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2605.12571