Synthetic Perception: Can Generated Images Unlock Latent Visual Prior for Text-Centric Reasoning?

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Huang, Yuesheng, Zhang, Peng, Wu, Xiaoxin, Liu, Riliang, Liang, Jiaqi
Format:	Preprint
Published:	2025
Subjects:	Multimedia Computer Vision and Pattern Recognition
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2506.17623