HyperAI超神経

プラットフォーム

3D Human Pose Estimation On Emdb

評価指標

Average MPJPE (mm)

Average MPJPE-PA (mm)

Average MVE (mm)

Average MVE-PA (mm)

Jitter (10m/s^3)

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

						Paper Title
PyMAF	131.065	82.8502	159.956	98.1305	81.8447	PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop
LGD	115.82	81.1389	140.63	95.7255	72.9695	Human Body Model Fitting by Learned Gradient Descent
FastMETRO-L	115.036	72.6765	133.566	86.0043	81.2959	Cross-Attention of Disentangled Modalities for 3D Human Mesh Recovery with Transformers
PARE	113.887	72.203	133.247	85.3788	75.1137	PARE: Part Attention Regressor for 3D Human Body Estimation
ROMP	112.652	75.1869	134.863	90.648	71.2556	Monocular, One-stage, Regression of Multiple 3D People
FastMETRO-L no SMPL Head	108.107	66.794	119.23	81.1847	185.933	Cross-Attention of Disentangled Modalities for 3D Human Mesh Recovery with Transformers
GLAMR	107.797	71.0091	128.237	85.4586	67.3522	GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras
CLIFF	103.134	68.7969	122.884	81.3275	55.4525	CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimation
HybrIK	103.037	65.5935	122.193	80.3678	49.2068	HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation
BioPose	92.5	52.1	98.9	-	-	BioPose: Biomechanically-accurate 3D Pose Estimation from Monocular Videos
GenHMR	88.2	51.7	99.5	-	-	GenHMR: Generative Human Mesh Recovery
WHAM (ViT)	79.7	50.4	94.4	-	-	WHAM: Reconstructing World-grounded Humans with Accurate 3D Motion
TRAM	74.4	45.7	86.6	-	-	TRAM: Global Trajectory and Motion of 3D Humans from in-the-wild Videos

0 of 13 row(s) selected.

3D Human Pose Estimation On Emdb | SOTA | HyperAI超神経