← 研踪 PaperTrailarXiv:2611.22495 · v1
Motion-Aware Reward Modeling for Instruction-Guided Video Editing
首次提交 2026-11-17 17:51
版本更新 2026-11-17 17:51
雷达观测 2026-11-18 11:33
摘要
Reward models for instruction-guided video editing can favor temporal consistency at the expense of motion. We study this failure, termed \emph{motion collapse}, using controlled motion-suppression variants of real edits. On VE-Bench, the evaluated composite reward prefers a motion-suppressed candidate in of cases. Learned video rewards and a VLM judge also struggle with mild suppression, while latent optimization and reward-guided fine-tuning exploit this weakness to increase reward as motion declines. A linear probe of our base reward backbone performs near chance on mild suppression, motivating an explicit inter-frame dynamics input. We propose a motion-aware reward model that combines an RGB frame-difference stream with preference training against motion-suppressed and fabricated-motion negatives. A distillation anchor preserves alignment with human quality judgments. On the VE-Bench motion ladder, our model reduces the rate of selecting motion-suppressed candidates to while achieving a Spearman correlation of with human quality ratings. It also detects unseen suppression variants and preserves motion during reward-guided fine-tuning on two editing pipelines. These results support explicitly modeling motion in rewards for video editing.
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-17 17:51 · 最新观测版本
Motion-Aware Reward Modeling for Instruction-Guided Video Editing
Reward models for instruction-guided video editing can favor temporal consistency at the expense of motion. We study this failure, termed \emph{motion collapse}, using controlled motion-suppression variants of real edits. On VE-Bench, the evaluated composite reward prefers a motion-suppressed candidate in of cases. Learned video rewards and a VLM judge also struggle with mild suppression, while latent optimization and reward-guided fine-tuning exploit this weakness to increase reward as motion declines. A linear probe of our base reward backbone performs near chance on mild suppression, motivating an explicit inter-frame dynamics input. We propose a motion-aware reward model that combines an RGB frame-difference stream with preference training against motion-suppressed and fabricated-motion negatives. A distillation anchor preserves alignment with human quality judgments. On the VE-Bench motion ladder, our model reduces the rate of selecting motion-suppressed candidates to while achieving a Spearman correlation of with human quality ratings. It also detects unseen suppression variants and preserves motion during reward-guided fine-tuning on two editing pipelines. These results support explicitly modeling motion in rewards for video editing.
查看此版本解读 →查看此版本原文 ↗