paper-with-me

GD-VCR

홈페이지 · 논문 6편

Geo-Diverse Visual Commonsense Reasoning (GD-VCR) is a new dataset to test vision-and-language models' ability to understand cultural and geo-location-specific commonsense. Image source: https://arxiv.org/pdf/2109.06860v1.pdf

ImagesTexts

벤치마크

Visual Commonsense Reasoning on GD-VCR 결과 4개