Bookmark and Share

Expanding Scene and Language Understanding with Large-scale Pre-training and a Unified Architecture (external)