AI RESEARCH
LibraGen: Playing a Balance Game in Subject-Driven Video Generation
arXiv CS.CV
•
ArXi:2603.13506v1 Announce Type: new With the advancement of video generation foundation models (VGFMs), customized generation, particularly subject-to-video (S2V), has attracted growing attention. However, a key challenge lies in balancing the intrinsic priors of a VGFM, such as motion coherence, visual aesthetics, and prompt alignment, with its newly derived S2V capability. Existing methods often neglect this balance by enhancing one aspect at the expense of others.