cs.CVNov 18, 2026

H3Remover: Reinforcing Video Object Removal without Paired Data

Authors: Kangjie Chen, Xiangyu Li, Jinhao Deng, Dongbin Zhang, Chaoda Zheng, Shijia Chen, Yuqi Zhang, Choo Sin Wai, 

Organizations: XPENG · THU · UCAS · CUHK

Abstract

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 edits such as recoloring, a target-removal reward favors calibrated depth changes consistent with revealing farther content. To limit unintended scene changes, a background-geometry reward encourages preservation of relative depth ordering. To preserve background appearance without retaining unwanted shadows or reflections, an appearance reward constrains RGB changes in source-derived protected regions while exempting targets and estimated effects. Frozen HPSv3 complements these constraints with perceptual-quality feedback for synthesized content. Evaluation regions remain fixed across candidates, and DiffusionART optimizes LoRA adapters with the backbone frozen. Experiments show substantial reconstruction gains over original H3, with qualitative comparisons demonstrating more effective object removal and better preservation of surrounding scene content.

Figures & tables

Explore similar work

CardsList
  1. TripleFlow: Training-Free Video Object Removal by Bridging Residual Editing and Native Generation

    Sep 30, 2026Songhe Wang, Lifu Wei, Shuolin Xu +2Video EditingImage Inpainting

  2. From Understanding to Erasing: Towards Complete and Stable Video Object Removal

    Apr 2, 2026Dingming Liu, Wenjing Wang, Chen Li +2Object RemovalVideo Editing

  3. VOID: Video Object and Interaction Deletion

    Apr 2, 2026Saman Motamed, William Harvey, Benjamin Klein +3Object RemovalVideo Editing