One-Forcing: 迈向稳定的一步自回归视频生成
One-Forcing: Towards Stable One-Step Autoregressive Video Generation
AI总结：
针对现有少步自回归视频生成方法在一步设置下质量严重下降的问题，提出One-Forcing方法，通过向DMD目标添加辅助GAN损失，实现高质量高效的一步视频生成，在VBench上达到83.76总分，成为一步因果视频生成中的最优方法。
摘要：
最近的进展显著改善了自回归机制下的实时交互式视频生成。然而，大多数现有的少步自回归视频生成方法（通常从相应的多步教师模型蒸馏而来）默认采用4步采样配置，这在部署期间仍会产生相当大的延迟，并且当进一步减少采样步数（特别是在一步设置中）时，会遭受严重的质量下降。轨迹式一致性蒸馏方法通常生成动态较弱的视频，而基于DMD的方法（如Self-Forcing）往往产生模糊的帧。为了解决这一挑战，我们提出了One-Forcing，一种简单而有效的方法，它通过向DMD目标添加辅助GAN损失，实现高质量高效的一步视频生成。在VBench上的实验表明，One-Forcing的总得分为83.76，在一步因果视频生成方法中达到了最先进的性能，并且与强大的多步方法保持竞争力。我们进一步证明，仅需分块模型三分之一的训练成本，即可稳定实现逐帧的一步自回归生成，而先前的方法未能成功实现这一设置。
英文摘要：
Recent advances in autoregressive diffusion-based video generation have substantially improved the quality of real-time video synthesis. However, most existing methods still require multiple denoising steps, while reducing sampling to a single step often leads to severe quality degradation: trajectory-based consistency distillation methods often produce videos with weak dynamics, whereas DMD-based methods, such as Self-Forcing, tend to generate blurry frames. We attribute this limitation to the teacher trajectories exhibiting highly concentrated curvature near the high-noise endpoint, which poses a fundamental geometric challenge for one-step distillation under the consistency distillation framework. To address these limitations, we propose One-Forcing, a simple yet effective approach that augments the DMD objective with an auxiliary GAN loss for high-quality and efficient one-step video generation. We further find that framewise autoregression stabilizes adversarial training, enabling higher-quality generation with substantially fewer training iterations than chunkwise autoregression. Experiments on VBench show that One-Forcing achieves a total score of 83.76, establishing state-of-the-art performance among one-step causal video generation methods while remaining competitive with strong multi-step approaches.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
