Story
arxiv_cs_ai · Nov 18, 2026 · paper
arxiv.orgNov 18, 2026
original source linked
In brief
Pretrained video editors possess rich removal and completion priors, yet may retain targets or alter unrelated content. We present H3Remover, which reinforces MiniMax-H3 Base Ref2VA for video object removal without paired object-free training targets. To discourage superficial ed