1 paper
Omer Gralnik, Guy Gafni, Ariel Shamir
We present Semantify: a self-supervised method that utilizes the semantic power of CLIP language-vision foundation model to simplify the control of 3D morphable models. Given a par…