研究人员开发了Steady-Forcing，一个旨在提高自回归扩散模型生成的长时域自然视频质量的新框架。该方法通过结合持久视觉锚点（V-Sink）和指数移动平均运动记忆（EMA-Sink）来解决场景布局漂移和运动抑制等常见问题。此外，该框架还纳入了块相对时间编码、周期性缓存净化以及从Wan2.1-14B教师模型进行蒸馏。评估表明，Steady-Forcing在扩展视频序列中增强了背景一致性和运动连续性，优于现有基线。 AI
影响 这项研究可能带来更稳定、更逼真的长格式视频生成，影响内容创作和模拟等应用。
排序理由 该集群包含两篇详细介绍视频生成新方法的学术论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2606.14732v1 Announce Type: cross Abstract: Autoregressive video diffusion models enable streaming generation but often degrade over long rollouts: static scene layouts drift, while mechanisms that improve spatial stability tend to suppress motion, causing natural flows suc…
Steady-Forcing addresses stability-motion trade-offs in long-horizon nature video generation through a memory and training framework combining visual anchors, motion memory, temporal encoding, and distillation techniques.
arXiv:2606.20310v1 Announce Type: new Abstract: Evaluating video generation with clean, pixel-based reward models disconnects evaluation from the noisy diffusion process and incurs massive VAE decoding costs. In this paper, we challenge this paradigm by asking a fundamental quest…
Evaluating video generation with clean, pixel-based reward models disconnects evaluation from the noisy diffusion process and incurs massive VAE decoding costs. In this paper, we challenge this paradigm by asking a fundamental question: Can a powerful video generator inherently d…