AI RESEARCH
Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation
arXiv CS.CV
•
ArXi:2606.04527v1 Announce Type: cross We present Echo Infinity, an autoregressive (AR) framework towards real-time infinite video generation that employs a learnable evolving memory to dynamically filter, abstract, and compress any-length history at constant cost. Existing methods mainly curate memory with predefined KV-cache schedules, fixed-ratio heuristic compression, or inference-time RoPE adaptation. These designs inevitably lose historical information and amplify compounding errors due to their limited cache window and ignorance of autoregressive generation noise.