Youming Zhang

DyCoDiff: Dynamic Collaborative Diffusion for Text‐Image‐Audio‐to‐Video

Text‐image‐audio‐to‐video (TIA2V) generation synthesizes videos whose semantics, appearance, and temporal dynamics are jointly controlled by text prompts, reference images, and audio signals. Existing TIA2V methods usually combine these modalities through separate spatial and tem...