StEP: Style-Based Encoder Pre-Training for Multi-Modal Image Synthesis.

Moustafa Meshry, Yixuan Ren, Larry S. Davis, Abhinav Shrivastava

Browse the full CVPR paper archive.