Home Console Docs News Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Human Instance Segmentation On Ochuman

Metrics

AP

Results

Performance results of various models on this benchmark

		Paper Title	Repository
HQNet (ViT-L)	38.8	You Only Learn One Query: Learning Unified Human Query for Single-Stage Multi-Person Multi-Task Human-Centric Perception
BBox-Mask-Pose 2x	32.4	Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing the Virtuous Circle
Crowd-SAM (ViT-L)	31.4	Crowd-SAM: SAM as a Smart Annotator for Object Detection in Crowded Scenes
HQNet (ResNet-50)	31.1	You Only Learn One Query: Learning Unified Human Query for Single-Stage Multi-Person Multi-Task Human-Centric Perception
BlendMask + CIS	29.8	Real-time Human-Centric Segmentation for Complex Video Scenes
Mask2Former + Occlusion C&P	28.3	Humans need not label more humans: Occlusion Copy & Paste for Occluded Human Instance Segmentation
CondInst + CIS	28.1	Real-time Human-Centric Segmentation for Complex Video Scenes
Mask2Former	27.8	Object-Centric Multi-Task Learning for Human Instances	-
HCQNet	27.3	Object-Centric Multi-Task Learning for Human Instances	-
ExPoSeg	26.8	PoSeg: Pose-Aware Refinement Network for Human Instance Segmentation	-
RTMDet-ins-l	26.5	Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing the Virtuous Circle
JoPoSeg	26.4	PoSeg: Pose-Aware Refinement Network for Human Instance Segmentation	-
BaseNet-DPS	25.5	Object-Centric Multi-Task Learning for Human Instances	-
Pose2Seg	23.8	Pose2Seg: Detection Free Human Instance Segmentation
PolarMask + CIS	23.4	Real-time Human-Centric Segmentation for Complex Video Scenes
ResNet-101-FPN + TTG v1	22.42	Test-time Adaptation vs. Training-time Generalization: A Case Study in Human Instance Segmentation using Keypoints Estimation	-
BCNet	20.6	Occlusion-Aware Instance Segmentation via BiLayer Network Architectures
CaSe	18.0	Count- and Similarity-aware R-CNN for Pedestrian Detection	-
Mask RCNN	16.9	Count- and Similarity-aware R-CNN for Pedestrian Detection	-

0 of 19 row(s) selected.

Human Instance Segmentation On Ochuman | SOTA | HyperAI