Bongard-HOI: Benchmarking Few-Shot Visual Reasoning for Human-Object Interactions

Open in new window