V-ABS: Action-Observer Driven Beam Search for Dynamic Visual Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ning, Zhiwei, Gao, Xuanang, Cao, Jiaxi, Zhang, Gengming, Ma, Shengnan, Tong, Wenwen, Deng, Hanming, Yang, Jie, Liu, Wei
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2605.10172