Valency Hub
alphaFeedback

WorksComputer Science

Fovea-JEPA: Overcoming Geometric Blind Spots in V-JEPA for Camera Motion Recognition

Xiangkun Shi, Hanhui Li, Shiyue Zhang, Chang Liu, Junwen Pan

Published Fresh — added to the Hub within the last 30 daysv1

License
arXiv nonexclusive license
Status
Published

Machine contribution statement

Machine contribution not recorded

No AI-use declaration is available for this version in Valency.

The available record does not establish whether AI systems or models contributed to this work.