Efficient T2V generation
Generate images from text prompts
Generate images from text descriptions
Create images from various types of annotations