TAG: Boosting Text-VQA via Text-aware Visual Question-answer Generation.

Published in: CoRR (2022)

Keyphrases