滚动扩散模型
Rolling Diffusion Models
AI总结：
提出 Rolling Diffusion，用滑动窗口去噪并为序列后期帧分配更多噪声，在 Kinetics-600 视频预测和混沌流体预测中优于标准扩散。
摘要：
扩散模型近来越来越多地应用于视频、流体力学模拟或气候数据等时序数据。这些方法通常在扩散过程中对后续帧施加等量噪声。本文探索了 Rolling Diffusion（滚动扩散）：一种采用滑动窗口去噪过程的新方法。它通过给序列中较晚出现的帧分配更多噪声，确保扩散过程随时间逐步破坏数据，从而反映生成过程展开时对未来更大的不确定性。实验表明，当时序动力学复杂时，Rolling Diffusion 优于标准扩散。具体而言，该结果在使用 Kinetics-600 视频数据集的视频预测任务以及混沌流体动力学预测实验中得到验证。
英文摘要：
Diffusion models have recently been increasingly applied to temporal data such as video, fluid mechanics simulations, or climate data. These methods generally treat subsequent frames equally regarding the amount of noise in the diffusion process. This paper explores Rolling Diffusion: a new approach that uses a sliding window denoising process. It ensures that the diffusion process progressively corrupts through time by assigning more noise to frames that appear later in a sequence, reflecting greater uncertainty about the future as the generation process unfolds. Empirically, we show that when the temporal dynamics are complex, Rolling Diffusion is superior to standard diffusion. In particular, this result is demonstrated in a video prediction task using the Kinetics-600 video dataset and in a chaotic fluid dynamics forecasting experiment.
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。
