StyleT2I: Toward Compositional and High-Fidelity Text-to-Image Synthesis.

Zhiheng Li, Martin Renqiang Min, Kai Li, Chenliang Xu

Browse the full CVPR paper archive.