Fetching the paper…

ERNIE-ViL: Knowledge Enhanced Vision-Language Representations Through Scene Graph · Around