Fetching the paper…

Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Models · Around