用双线性目标直接生成视频，速度快且误差少。
Generative Video Bi-flow
- 将过去帧直接映射到未来帧，跳过噪声中间步骤。
- 训练时加噪声，有效减少累积误差，提升生成稳定性。
- 无需条件输入，实时流式生成，速度优于扩散模型。
我们提出一种新型生成视频模型，通过神经常微分方程（ODE）流来稳健学习时间变化。该模型采用双线性目标，一方面直接从历史帧生成未来帧，避免传统方法中从噪声生成新帧的高计算开销；另一方面，通过训练时添加噪声，联合学习消除累积误差。我们在多个视频数据集上实现了无条件、流式视频生成，在生成质量上与条件扩散基线相当，但仅需更少的ODE求解步数，显著提升生成速度。
原文摘要 · Abstract (English)
We propose a novel generative video model to robustly learn temporal change as a neural Ordinary Differential Equation (ODE) flow with a bilinear objective which combines two aspects: The first is to map from the past into future video frames directly. Previous work has mapped the noise to new frames, a more computationally expensive process. Unfortunately, starting from the previous frame, instead of noise, is more prone to drifting errors. Hence, second, we additionally learn how to remove the accumulated errors as the joint objective by adding noise during training. We demonstrate unconditional video generation in a streaming manner for various video datasets, all at competitive quality compared to a conditional diffusion baseline but with higher speed, i.e., fewer ODE solver steps.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。