Great work! Some questions

[fxyang9]

Thanks for the awesome work! Really amazing generation results! It seems some related works like [1,2], which employs bi-directional transformer for generation, are not mentioned or compared in the paper. I am a little curious about how the performance differs.

[1] High-Fidelity Pluralistic Image Completion with Transformers, ICCV 2021
[2] M6-UFC: Unifying Multi-Modal Controls for Conditional Image Synthesis, Arxiv 2021