Exposing the Limits of Video-Text Models through Contrast Sets.

Jae Sung Park, Sheng Shen, Ali Farhadi, Trevor Darrell, Yejin Choi, Anna Rohrbach

Browse the full NAACL paper archive.