Unifying Vision-and-Language Tasks via Text Generation.

Jaemin Cho, Jie Lei, Hao Tan, Mohit Bansal

Browse the full ICML paper archive.