AI RESEARCH
FrameDiT: Diffusion Transformer with Frame-Level Matrix Attention for Efficient Video Generation
arXiv CS.CV
•
ArXi:2603.09721v1 Announce Type: new High-fidelity video generation remains challenging for diffusion models due to the difficulty of modeling complex spatio-temporal dynamics efficiently. Recent video diffusion methods typically represent a video as a sequence of spatio-temporal tokens which can be modeled using Diffusion Transformers (DiTs). However, this approach faces a trade-off between the strong but expensive Full 3D Attention and the efficient but temporally limited Local Factorized Attention.