Papers2

#high-resolution video generation

Efficient Autoregressive Video Diffusion with Dummy Head

Hang Guo, Zhaoyang Jia et al.Jan 28arXiv

This paper finds that about 1 out of every 4 attention heads in autoregressive video diffusion models mostly looks only at the current frame and almost ignores the past, wasting memory and time.

#autoregressive video diffusion#multi-head self-attention#KV cache compression

HiStream: Efficient High-Resolution Video Generation via Redundancy-Eliminated Streaming

Intermediate

Haonan Qiu, Shikun Liu et al.Dec 24arXiv

HiStream makes 1080p video generation much faster by removing repeated work across space, time, and steps.

#high-resolution video generation#diffusion transformer (DiT)#dual-resolution caching