Unpaired Image-to-Speech Synthesis With Multimodal Information Bottleneck.

Shuang Ma, Daniel McDuff, Yale Song

Browse the full ICCV paper archive.