Alibaba's open-source video generation model Wan2.2-S2V

On August 26th, it was reported that Alibaba has open-sourced a brand-new multimodal video generation model, Tongyi Wan2.2-S2V. With just one static image and a piece of audio, it can generate movie-level digital human videos with natural facial expressions, consistent mouth movements, and smooth body movements. The video duration generated by this model in a single attempt can reach the minute level. Significantly enhance the video creation efficiency in industries such as digital human live streaming, film and television production, and AI education

.