# Mask Forcing：通过双噪声掩码滚动提升自回归视频扩散蒸馏质量

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-09-08 08:00
- AIHOT 分数：36
- AIHOT 链接：https://aihot.news/items/cmttmozwp0ejdrofpm5agyhcn
- 原文链接：https://arxiv.org/abs/2609.09123

## AI 摘要

针对自回归视频扩散模型在分布匹配蒸馏中因逆向KL目标导致模式坍缩、生成视频过饱和与过平滑的问题，研究者提出Mask Forcing策略。该方法在自回归学生模型自滚动过程中沿空间与时间轴注入随机掩码，以引入更干净的信号并缓解模式坍缩。实验表明，该方法能高效提升多种自回归视频扩散蒸馏方法的视觉质量，且无需真实视频数据或额外后训练阶段。

## 正文

Autoregressive (AR) video diffusion models have shown great potential in real-time video generation. Recent methods distill pretrained bidirectional video diffusion models into causal AR students through Distribution Matching Distillation (DMD), but the generated videos often suffer from over-saturation and over-smoothing issues, resulting in limited visual quality and realism. The key contributing factor is the mode-seeking behavior of the reverse KL objective in DMD, which can cause the student distribution to collapse onto only a few modes of the teacher distribution. To address this, we propose Mask Forcing, a Dual-Noise Masking Rollout strategy that perturbs the AR student self-rollout to mitigate mode collapse induced by reverse-KL mode seeking. The core idea is to inject cleaner signals into noisy rollout inputs via random masks along spatial and temporal axes during the self-rollout process of AR diffusion distillation. Such perturbations encourage the student rollouts to explore more regions of the teacher distribution, allowing DMD to provide learning signals beyond the modes already covered by the student. Moreover, the cleaner tokens act as denoising guidance for other noisier tokens, improving the intermediate rollout predictions and reducing error accumulation. Extensive experiments demonstrate that our method improves multiple AR video diffusion distillation methods with higher visual quality efficiently, without incorporating real video data or additional post-training stages.