See Further, Think Deeper: Advancing VLM's Reasoning Ability with Low-level Visual Cues and Reflection

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Zhiheng, Wang, Tong, Wang, Shuning, Liu, Naiming, Zhang, Yumeng
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/2604.24339