Text-to-video generation has lagged behind text-to-image synthesis in quality and diversity
due to the complexity of spatio-temporal modeling and limited video-text datasets. This …