Action Recognition with Kernel-based Graph Convolutional Networks
This work addresses the limitations of existing GCN aggregation schemes for researchers and practitioners working on graph-based deep learning, offering a more robust and efficient convolutional approach.
This paper introduces a novel Graph Convolutional Network (GCN) framework that performs spatial graph convolution in a Reproducing Kernel Hilbert Space (RKHS). This approach allows for the design of convolutional graph filters in a high-dimensional, more discriminating space without increasing training parameters, and achieves superior performance on skeleton-based action recognition compared to existing methods.
Learning graph convolutional networks (GCNs) is an emerging field which aims at generalizing deep learning to arbitrary non-regular domains. Most of the existing GCNs follow a neighborhood aggregation scheme, where the representation of a node is recursively obtained by aggregating its neighboring node representations using averaging or sorting operations. However, these operations are either ill-posed or weak to be discriminant or increase the number of training parameters and thereby the computational complexity and the risk of overfitting. In this paper, we introduce a novel GCN framework that achieves spatial graph convolution in a reproducing kernel Hilbert space (RKHS). The latter makes it possible to design, via implicit kernel representations, convolutional graph filters in a high dimensional and more discriminating space without increasing the number of training parameters. The particularity of our GCN model also resides in its ability to achieve convolutions without explicitly realigning nodes in the receptive fields of the learned graph filters with those of the input graphs, thereby making convolutions permutation agnostic and well defined. Experiments conducted on the challenging task of skeleton-based action recognition show the superiority of the proposed method against different baselines as well as the related work.