Self Supervised Human Action Recognition On

Metrics

Classifier

Encoder

xset (%)

xsub (%)

Results

Performance results of various models on this benchmark

					Paper Title
AS-CAL	FC	LSTM	49.2	48.6	Augmented Skeleton Based Contrastive Action Learning with Momentum LSTM for Unsupervised Action Recognition
3s-AimCLR	FC	ST-GCN	68.8	68.2	Contrastive Learning from Extremely Augmented Skeleton Sequences for Self-supervised Action Recognition
3s-RVTCLR+	FC	ST-GCN	68.9	68.0	Modeling the Relative Visual Tempo for Self-supervised Skeleton-based Action Recognition
PCRP	FC	GRU	45.1	41.7	Prototypical Contrast and Reverse Prediction: Unsupervised Skeleton Based Action Recognition
P&C	FC	GRU	42.7	41.7	PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition
MCAE	FC	MCAE	54.7	52.8	Unsupervised Motion Representation Learning with Capsule Autoencoders
CMCS	FC	ST-GCN	71.7	68.5	Cross-Model Cross-Stream Learning for Self-Supervised Human Action Recognition
SDS-CL	FC	DSTA	55.6	50.6	Spatiotemporal Decouple-and-Squeeze Contrastive Learning for Semi-Supervised Skeleton-based Action Recognition

0 of 8 row(s) selected.

Self Supervised Human Action Recognition On

Metrics

Classifier

Encoder

xset (%)

xsub (%)

Results

Performance results of various models on this benchmark

					Paper Title
AS-CAL	FC	LSTM	49.2	48.6	Augmented Skeleton Based Contrastive Action Learning with Momentum LSTM for Unsupervised Action Recognition
3s-AimCLR	FC	ST-GCN	68.8	68.2	Contrastive Learning from Extremely Augmented Skeleton Sequences for Self-supervised Action Recognition
3s-RVTCLR+	FC	ST-GCN	68.9	68.0	Modeling the Relative Visual Tempo for Self-supervised Skeleton-based Action Recognition
PCRP	FC	GRU	45.1	41.7	Prototypical Contrast and Reverse Prediction: Unsupervised Skeleton Based Action Recognition
P&C	FC	GRU	42.7	41.7	PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition
MCAE	FC	MCAE	54.7	52.8	Unsupervised Motion Representation Learning with Capsule Autoencoders
CMCS	FC	ST-GCN	71.7	68.5	Cross-Model Cross-Stream Learning for Self-Supervised Human Action Recognition
SDS-CL	FC	DSTA	55.6	50.6	Spatiotemporal Decouple-and-Squeeze Contrastive Learning for Semi-Supervised Skeleton-based Action Recognition

0 of 8 row(s) selected.