DreamBooth

: Google AI Research DreamBooth is a new tool that utilizes text-to-image technology to personalize images of a subject using only 3-5 user provided photos. This is accomplished by embedding a unique identifier into the output domain of the model, which can then be used to generate realistic images contextualized in different scenes. DreamBooth was developed by Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein and Kfir Aberman – from Google AI Research. With this tool users can create photorealistic images with specified viewpoint for their subjects as well as synthesizing novel pictures from existing ones.