Flowception：面向视频生成的时间可扩展流匹配
Flowception: Temporally Expansive Flow Matching for Video Generation
把「离散插帧 + 连续去噪」交替的概率路径用于视频生成，可作为一种非自回归、可变长度的采样范式，替代自回归视频模型以缓解误差累积，并天然支持图像到视频与插帧等任务。
原文依据 Flowception learns a probability path that interleaves discrete frame insertions with continuous frame denoising.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 Flowception，一种新颖的非自回归、可变长度视频生成框架。Flowception 学习一条在离散帧插入与连续帧去噪之间交替的概率路径。与自回归方法相比，Flowception 缓解了误差累积与漂移，因为采样过程中的帧插入机制作为一种高效的压缩机制来处理长程上下文。与全序列流方法相比，本方法将训练的浮点运算量降低至三分之一，同时更适配局部注意力变体，并允许将视频长度与其内容联合学习。定量实验结果表明，相较于自回归与全序列基线，本方法在 FVD 与 VBench 指标上均有提升，定性结果进一步验证了这一点。最后，通过学习在序列中插入并去噪帧，Flowception 无缝整合了图像到视频生成与视频插帧等不同任务。
英文原文摘要 980 字符
We present Flowception, a novel non-autoregressive and variable-length video generation framework. Flowception learns a probability path that interleaves discrete frame insertions with continuous frame denoising. Compared to autoregressive methods, Flowception alleviates error accumulation/drift as the frame insertion mechanism during sampling serves as an efficient compression mechanism to handle long-term context. Compared to full-sequence flows, our method reduces FLOPs for training three-fold, while also being more amenable to local attention variants, and allowing to learn the length of videos jointly with their content. Quantitative experimental results show improved FVD and VBench metrics over autoregressive and full-sequence baselines, which is further validated with qualitative results. Finally, by learning to insert and denoise frames in a sequence, Flowception seamlessly integrates different tasks such as image-to-video generation and video interpolation.