DynFrame: Adaptive Reasoning-Driven Multimodal Framework with Dynamic Frame Augmentation for Complex Video Understanding

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhang, Peng, Zhang, Guanghao, He, Wanggui, Zhang, Longxiang, Liu, Mushui, Xia, Yan, Peng, Zhenhao, Dai, Weilong, Liu, Jinlong, Tang, Haobing, Zhang, Le, Jiang, Hao, Huang, Pipei
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/2605.26680