PulseAugur
实时 07:35:09
English(EN) FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching

FlowLong 支持在不重新训练的情况下生成更长的视频

研究人员开发了FlowLong,一种新颖的推理时方法,用于扩展视频扩散模型生成更长序列的能力。该方法使用重叠的滑动窗口和一种称为Tweedie匹配的技术,以确保时间一致性并保持视觉质量,而无需额外的训练。FlowLong不依赖于特定架构,并已成功地延长了视频生成长度,同时也适用于音视频联合生成和文本到3D场景生成。 AI

影响 使扩散模型能够在不进行额外训练的情况下生成更长、更一致的视频。

排序理由 该集群包含一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 2 个来源。 我们如何撰写摘要 →

报道来源 [2]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching

    Extending the generation horizon of video diffusion models to long sequences remains a long-standing and important challenge. Existing training-free approaches fall into two categories: extensions of bidirectional models, which are tightly coupled to specific architectures and su…

  2. Hugging Face Daily Papers TIER_1 English(EN) ·

    FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching

    A novel inference-time method for long video generation using overlapping sliding windows with Tweedie matching and stochastic early-phase sampling to improve temporal consistency and visual quality.