Graph convolutional networks (GCNs) are an effective skeleton-based human action recognition (HAR) technique. GCNs enable the specification of CNNs to a non-Euclidean frame that is more flexible. The previous GCN-based models still have a lot of issues: (I) The graph structure is the same for all model layers and input data.
@article{arxiv.2411.06553,
title = {Extended multi-stream temporal-attention module for skeleton-based human action recognition (HAR)},
author = {Faisal Mehmood and Xin Guo and Enqing Chen and Muhammad Azeem Akbar and Arif Ali Khan and Sami Ullah},
journal= {arXiv preprint arXiv:2411.06553},
year = {2024}
}
Comments
This paper accepted in Computers in Human Behavior Journal