ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks

Date:

At Vision and Language seminar, I presented the following paper and discussed its pros and cons!

ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks

  • Jiasen Lu, Dhruv Batra, Devi Parikh, Stefan Lee and its criticism.