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

English

3D Human Pose Estimation On Agora

Metrics

B-MPJPE

B-MVE

B-NMJE

B-NMVE

Results

Performance results of various models on this benchmark

Model Name	B-MPJPE	B-MVE	B-NMJE	B-NMVE	Paper Title	Repository
Multi-Person 3D Pose and Shape Estimationvia Inverse Kinematics and Refinement	91.6	86.7	110.4	104.5	Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement
HybrIK	77	73.9	84.6	81.2	HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-body Mesh Recovery
NIKI (Twist-and-Swing)	67.3	63.9	74	70.2	NIKI: Neural Inverse Kinematics with Invertible Neural Networks for 3D Human Pose and Shape Estimation
PARE	146.2	140.9	174.0	167.7	PARE: Part Attention Regressor for 3D Human Body Estimation
HMR	180.5	173.6	226.0	217.0	End-to-end Recovery of Human Shape and Pose
HybrIK-X	67.2	68.5	72.3	73.7	HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-body Mesh Recovery
SPEC	112.3	106.5	133.7	126.8	SPEC: Seeing People in the Wild with an Estimated Camera
W-HMR	67.9	63.4	75.4	70.4	W-HMR: Monocular Human Mesh Recovery in World Space with Weak-Supervised Calibration
PyMAF	83.3	78.6	92.6	87.3	PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop
PyMAF-X	83.2	84.0	93.5	94.4	PyMAF-X: Towards Well-aligned Full-body Model Regression from Monocular Images
SPIN	175.1	168.7	223.1	216.3	Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop

0 of 11 row(s) selected.