Connect, Collapse, Corrupt: Learning Cross-Modal Tasks with Uni-Modal Data.

Yuhui Zhang, Elaine Sui, Serena Yeung

Browse the full ICLR paper archive.