通过动态帧选择与云边协同，实现低延迟视频生成。
EcoVideo: Entropy-Orchestrated Video Generation Paradigm in Cloud-Edge Dynamics
- 基于自注意力熵自动筛选关键帧，无需额外训练。
- 在低带宽下实现最高2.9倍的端到端加速，保持高质量输出。
- 适合资源受限的边缘设备，支持实时系统动态调整。
DiT视频生成因迭代全帧去噪而延迟高，现有云边方法多依赖静态跨步解耦，无法利用帧间相似性或适应系统动态。我们提出EcoVideo，一种熵驱动的动态帧间解耦框架：早期自注意力熵提供无需训练的帧级信息密度估计，用于帧选择；云端大模型对稀疏的高熵关键帧进行去噪；边缘轻量模型通过运动感知插值并优化重构剩余帧，保障时序稳定性。EcoVideo进一步根据实时带宽和算力动态调整关键帧预算与边缘优化深度，在端到端延迟约束下实现最优性能。在典型DiT视频生成器上的实验表明，该方法显著提升质量-效率权衡，在低带宽、算力受限的边缘场景下实现最高2.9倍的端到端加速。代码已开源：https://github.com/IF-LAB-PKU/EcoVideo。
原文摘要 · Abstract (English)
DiT video generation is latency-intensive due to iterative full-frame denoising, while prior cloud-edge methods largely rely on static inter-step decoupling and cannot leverage inter-frame similarity or adapt to system dynamics. We propose EcoVideo, an entropy-orchestrated framework for dynamic inter-frame decoupling: early-stage self-attention entropy provides a training-free estimate of frame-wise information density for frame selection; a cloud large model denoises sparse high-entropy keyframes; and an edge lightweight model reconstructs the remaining frames via motion-aware interpolation with refinement for temporal stability. EcoVideo further adapts the keyframe budget and edge refinement depth to real-time bandwidth and compute availability, optimizing end-to-end latency under constraints. Experiments on representative DiT video generators show improved quality--efficiency trade-offs and up to 2.9x end-to-end speedup in low-bandwidth, compute-limited edge settings. Code is available at https://github.com/IF-LAB-PKU/EcoVideo.
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。