ALEXANDRIA, Va., Feb. 24 -- United States Patent no. 12,561,857, issued on Feb. 24, was assigned to ADOBE INC. (San Jose, Calif.).
"Multi-attribute inversion for text-to-image synthesis" was invented by Aishwarya Agarwal (Karnataka, India), Srikrishna Karanam (Karnataka, India), Tripti Shukla (Karnataka, India) and Balaji Vasan Srinivasan (Karnataka, India).
According to the abstract* released by the U.S. Patent & Trademark Office: "An image generation model obtains a text prompt, a first attribute token, and a second attribute token. A first set of layers of the image generation model and a first set of time-steps are identified for the first attribute token and a second set of layers of the image generation model and a second set of tim...