1 paper · 1 filter
Nokyung Park, Daewon Chae, Jeongyong Shim +3
Learning domain-invariant visual representations is important to train a model that can generalize well to unseen target task domains. Recent works demonstrate that text descriptio…