VQA-MHUG

홈페이지 · 논문 1편

VQA-MHUG is a 49-participant dataset of multimodal human gaze on both images and questions during visual question answering (VQA) collected using a high-speed eye tracker.