Learning Visual Affordance Grounding from Demonstration Videos

Benchmark Model Rank Results
video-to-image-affordance-grounding-on-epic-hotspotHAG-Net (+Hand Box)–KLD: 1.21SIM: 0.41AUC-J: 0.80
video-to-image-affordance-grounding-on-opra-28x28HAG-Net (+Hand Box)–KLD: 1.41SIM: 0.37AUC-J: 0.81