Skeleton Based Action Recognition On J Hmbd

10%

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	10%
relational-autoencoder-for-feature-extraction	60.6
graph-attention-networks	58.1