UniRAP: Towards Unified Part-level Physical Affordance Reasoning and Actionable Perception

Benchmark Model Rank Results
referring-video-object-segmentation-on-revosUniRAP–J&F: 61.9J: 59.3F: 64.4