-
EndoVis-17-VQLA
EndoVis-17-VQLA dataset is a public dataset with 97 frames with common tools and interactions from EndoVis-2017. It is annotated with question-answer-bounding box labels. -
EndoVis-18-VQLA
EndoVis-18-VQLA dataset is a public dataset with 14 video sequences on robotics surgery procedures. It is combined with the bounding box on tissue-instrument interaction...