Video Diffusion Transformer

A video generation architecture that combines the stepwise denoising process of diffusion models with the self-attention mechanism of Transformers, enabling flexible handling of long sequences and cross-modal conditioning.