Imagen Text to Video AI

Tech companies all over the world are competing in developing more advanced AI generators. Google just presented their latest proposal - Imagen Video - an AI system that creates videos from text prompts

From a written instruction the system generates and upscales to 128 image frames of video with 24 frames per second in 720p resolution. As with text to image generators Google plans to have it e.g. imitate artistic styles. In its current state it is considered not yet mature enough to be released to the public. Imagen Video builds on Google’s Imagen, an image-generating system comparable to OpenAI’s DALL-E 2 and Stable Diffusion.  An image to video was released from Meta last last week. 

Research paper