Tasks
SotA
Datasets
Papers
Methods
Submit
About
SotA
/
Computer Vision
/
Pose Estimation
/
Human3.6M
Pose Estimation on Human3.6M
Metric: Average MPJPE (mm) @ 400ms (higher is better)
Leaderboard
Dataset
Loading chart...
Results
Submit a result
Export CSV
Sort:
Average MPJPE (mm) @ 400ms (best first)
Average MPJPE (mm) @ 400ms (worst first)
Date (newest first)
Date (oldest first)
Model name (A→Z)
#
Model
↕
Average MPJPE (mm) @ 400ms
▼
Extra Data
Paper
Date
↕
Code
1
Forecast LSTM
80.8
No
-
-
Code
2
AnyPose1
80.6
No
AnyPose: Anytime 3D Human Pose Forecasting via N...
2023-09-09
-
3
STS-GCN
65.8
No
Space-Time-Separable Graph Convolutional Network...
2021-10-09
Code
4
LTD-GCN
63.5
No
Learning Trajectory Dependencies for Human Motio...
2019-08-15
Code
5
MSR-GCN
62.9
No
MSR-GCN: Multi-Scale Residual Graph Convolution ...
2021-08-16
Code
6
UNIMASK-M
61.6
No
A Unified Masked Autoencoder with Patchified Ske...
2023-08-14
-
7
MotionMixer
59.3
No
MotionMixer: MLP-based 3D Human Body Pose Foreca...
2022-07-01
Code
8
PGBIG
58.5
No
Progressively Generating Better Initial Guesses ...
2022-03-30
Code
9
siMLPe
57.3
No
Back to MLP: A Simple Baseline for Human Motion ...
2022-07-04
Code
10
GraphMixer
56.7
No
Graph-Guided MLP-Mixer for Skeleton-Based Human ...
2023-04-07
-
11
ResChunk
56.41
No
Multiscale Residual Learning of Graph Convolutio...
2023-08-31
Code
12
Sun et al.
55.5
No
Towards Accurate Human Motion Prediction via Ite...
2023-05-08
-
13
EqMotion
55
No
EqMotion: Equivariant Multi-agent Motion Predict...
2023-03-20
Code
14
AuxFormer
54.1
No
Auxiliary Tasks Benefit 3D Skeleton-based Human ...
2023-08-17
Code
15
FMS-AM
34.8
No
Remembering What Is Important: A Factorised Mult...
2023-05-19
-
16
MGCN
34.4
No
Multi-Graph Convolution Network for Pose Forecas...
2023-04-11
-
17
GCNext
30.5
No
GCNext: Towards the Unity of Graph Convolutions ...
2023-12-19
Code
#1
Forecast LSTM
80.8
Average MPJPE (mm) @ 400ms
No paper
Code
#2
AnyPose1
SOTA
80.6
Average MPJPE (mm) @ 400ms
· 2023-09-09
AnyPose: Anytime 3D Human Pose Forecasting via Neural Ordinary Differential Equations
#3
STS-GCN
SOTA
65.8
Average MPJPE (mm) @ 400ms
· 2021-10-09
Space-Time-Separable Graph Convolutional Network for Pose Forecasting
Code
#4
LTD-GCN
SOTA
63.5
Average MPJPE (mm) @ 400ms
· 2019-08-15
Learning Trajectory Dependencies for Human Motion Prediction
Code
#5
MSR-GCN
62.9
Average MPJPE (mm) @ 400ms
· 2021-08-16
MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human Motion Prediction
Code
#6
UNIMASK-M
61.6
Average MPJPE (mm) @ 400ms
· 2023-08-14
A Unified Masked Autoencoder with Patchified Skeletons for Motion Synthesis
#7
MotionMixer
59.3
Average MPJPE (mm) @ 400ms
· 2022-07-01
MotionMixer: MLP-based 3D Human Body Pose Forecasting
Code
#8
PGBIG
58.5
Average MPJPE (mm) @ 400ms
· 2022-03-30
Progressively Generating Better Initial Guesses Towards Next Stages for High-Quality Human Motion Prediction
Code
#9
siMLPe
57.3
Average MPJPE (mm) @ 400ms
· 2022-07-04
Back to MLP: A Simple Baseline for Human Motion Prediction
Code
#10
GraphMixer
56.7
Average MPJPE (mm) @ 400ms
· 2023-04-07
Graph-Guided MLP-Mixer for Skeleton-Based Human Motion Prediction
#11
ResChunk
56.41
Average MPJPE (mm) @ 400ms
· 2023-08-31
Multiscale Residual Learning of Graph Convolutional Sequence Chunks for Human Motion Prediction
Code
#12
Sun et al.
55.5
Average MPJPE (mm) @ 400ms
· 2023-05-08
Towards Accurate Human Motion Prediction via Iterative Refinement
#13
EqMotion
55
Average MPJPE (mm) @ 400ms
· 2023-03-20
EqMotion: Equivariant Multi-agent Motion Prediction with Invariant Interaction Reasoning
Code
#14
AuxFormer
54.1
Average MPJPE (mm) @ 400ms
· 2023-08-17
Auxiliary Tasks Benefit 3D Skeleton-based Human Motion Prediction
Code
#15
FMS-AM
34.8
Average MPJPE (mm) @ 400ms
· 2023-05-19
Remembering What Is Important: A Factorised Multi-Head Retrieval and Auxiliary Memory Stabilisation Scheme for Human Motion Prediction
#16
MGCN
34.4
Average MPJPE (mm) @ 400ms
· 2023-04-11
Multi-Graph Convolution Network for Pose Forecasting
#17
GCNext
30.5
Average MPJPE (mm) @ 400ms
· 2023-12-19
GCNext: Towards the Unity of Graph Convolutions for Human Motion Prediction
Code