You Only Learn One Query: Learning Unified Human Query for Single-Stage Multi-Person Multi-Task Human-Centric Perception

Open paper
Benchmark Model Rank Results
human-instance-segmentation-on-ochumanHQNet (ResNet-50)#3AP: 31.1
pose-estimation-on-ochumanHQNet (ViT-L)#7Test AP: 45.6
pose-estimation-on-ochumanHQNet (ResNet-50)#11Test AP: 40.0

Rank counts only results with a code link.