Skeleton Based Action Recognition On Ntu Rgbd

评估指标

Accuracy (CS)

Accuracy (CV)

评测结果

各个模型在此基准测试上的表现结果

模型名称	Accuracy (CS)	Accuracy (CV)	Paper Title	Repository
ST-GCN [PYSKL, 3D Skeleton]	90.7	96.5	Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition
H-RNN	59.1	64.0	Hierarchical recurrent neural network for skeleton based action recognition	-
Action Capsules	90	96.3	Action Capsules: Human Skeleton Action Recognition	-
3s-ActCLR	84.3	88.8	Actionlet-Dependent Contrastive Learning for Unsupervised Skeleton-Based Action Recognition	-
MS-AAGCN+TEM	91.0	96.5	Temporal Extension Module for Skeleton-Based Action Recognition	-
IndRNN (with jpd)	83	89	A Comparative Review of Recent Kinect-based Action Recognition Algorithms
EleAtt-GRU (aug.)	80.7	88.4	EleAtt-RNN: Adding Attentiveness to Neurons in Recurrent Neural Networks	-
Ind-RNN	81.8	88.0	Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN
CoAGCN* (2-stream)	86.0	93.1	Continual Spatio-Temporal Graph Convolutional Networks
CoST-GCN* (2-stream)	88.3	95	Continual Spatio-Temporal Graph Convolutional Networks
DualHead-Net	92.0	96.6	Learning Multi-Granular Spatio-Temporal Graph Network for Skeleton-based Action Recognition
Clips+CNN+MTLN	79.6	84.8	A New Representation of Skeleton Sequences for 3D Action Recognition	-
Hyperformer	92.9	96.5	Hypergraph Transformer for Skeleton-based Action Recognition
FO-GASTM	82.83	90.05	Learning Shape-Motion Representations from Geometric Algebra Spatio-Temporal Model for Skeleton-Based Action Recognition	-
2s-NLGCN	88.5	95.1	Non-Local Graph Convolutional Networks for Skeleton-Based Action Recognition
CTR-GCN	92.4	96.8	Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition
Spatio-Temporal LSTM	69.2	77.7	Spatio-Temporal LSTM with Trust Gates for 3D Human Action Recognition	-
CNN+Motion+Trans	83.2	89.3	Skeleton-based Action Recognition with Convolutional Neural Networks
3s-HYSP	89.1	95.2	HYperbolic Self-Paced Learning for Self-Supervised Skeleton-based Action Representations
CoS-TR*	86.3	92.4	Continual Spatio-Temporal Graph Convolutional Networks

0 of 133 row(s) selected.