Self-Organizing Kernel-based Convolutional Echo State Network for Human Actions Recognition

We propose a deterministic initialization of the Echo State Network reservoirs to ensure that the activation of its internal echo state representations reflects similar topological qualities of the input signal which should lead to a self-organizing reservoir. Human actions encoded as a multivariate time series signal are clustered before using the clustered nodes and interconnectivity matrices for initializing the S-ConvESN reservoirs. The capability of S-ConvESN is evaluated using several 3D-skeleton-based action recognition datasets. <br><br> <a href="ES2020-SUN_12_07_99_LOO_poster.mp4" target="blank">Poster video</a> <br> <a href="ES2020-SUN_12_07_99_LOO_spotlight.mp4" target="blank">Spotlight video</a>

@InProceedings{LLLW20, 
 	 author =  {Lee, Gin Chong and Loo, Chu Kiong and Liew, Wei Shiung and Wermter, Stefan},  
 	 title = {Self-Organizing Kernel-based Convolutional Echo State Network for Human Actions Recognition}, 
 	 booktitle = {Proceedings of the European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN’20)},
 	 number = {},
 	 volume = {},
 	 pages = {591--596},
 	 year = {2020},
 	 month = {},
 	 publisher = {},
 	 doi = {}, 
 }