HyperAI超神经

首页算力平台文档资讯论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

2D Human Pose Estimation On Human Art

评估指标

AP

AP (gt bbox)

评测结果

各个模型在此基准测试上的表现结果

			Paper Title	Repository
UniPose	0.759	-	X-Pose: Detecting Any Keypoints
ED-Pose (R50)	0.723	/	Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimation
ViTPose-h	0.468	0.800	ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation
ViTPose-l	0.459	0.789	ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation
HRNet-w48	0.417	0.769	Deep High-Resolution Representation Learning for Human Pose Estimation
ViTpose-b	0.410	0.759	ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation
HRNet-w32	0.399	0.754	Deep High-Resolution Representation Learning for Human Pose Estimation
ViTPose-s	0.381	0.738	ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation
RTMPose-s	0.311	-	RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose
RTMPose-l	-	0.753	RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose

0 of 10 row(s) selected.

2D Human Pose Estimation On Human Art | SOTA | HyperAI超神经