Learning Chebyshev Basis in Graph Convolutional Networks for Skeleton-based Action Recognition
arXiv:2104.05482
Abstract
Spectral graph convolutional networks (GCNs) are particular deep models which aim at extending neural networks to arbitrary irregular domains. The principle of these networks consists in projecting graph signals using the eigen-decomposition of their Laplacians, then achieving filtering in the spectral domain prior to back-project the resulting filtered signals onto the input graph domain. However, the success of these operations is highly dependent on the relevance of the used Laplacians which are mostly handcrafted and this makes GCNs clearly sub-optimal. In this paper, we introduce a novel spectral GCN that learns not only the usual convolutional parameters but also the Laplacian operators. The latter are designed "end-to-end" as a part of a recursive Chebyshev decomposition with the particularity of conveying both the differential and the non-differential properties of the learned representations -- with increasing order and discrimination power -- without overparametrizing the trained GCNs. Extensive experiments, conducted on the challenging task of skeleton-based action recognition, show the generalization ability and the outperformance of our proposed Laplacian design w.r.t. different baselines (built upon handcrafted and other learned Laplacians) as well as the related work.
References in corpus (23)
- Adam: A Method for Stochastic Optimization
- Semi-Supervised Classification with Graph Convolutional Networks
- Deep Residual Learning for Image Recognition
- The Emerging Field of Signal Processing on Graphs: Extending High-Dimensional Data Analysis to Networks and Other Irregular Domains
- Inductive Representation Learning on Large Graphs
- Spectral Networks and Locally Connected Networks on Graphs
- Deep Convolutional Networks on Graph-Structured Data
- Learning Combinatorial Optimization Algorithms over Graphs
- Learning Convolutional Neural Networks for Graphs
- Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition
- Diffusion-Convolutional Neural Networks
- FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling
- Learning graphs from data: A signal representation perspective
- NTU RGB+D: A Large Scale Dataset for 3D Human Activity Analysis
- GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphs
- Adaptive Sampling Towards Fast Graph Representation Learning
- Graph topology inference based on sparsifying transform learning
- View Adaptive Recurrent Neural Networks for High Performance Human Action Recognition from Skeleton Data
- Deep Iterative and Adaptive Learning for Graph Neural Networks
- Geometric deep learning on graphs and manifolds using mixture model CNNs
- Modular meta-learning
- Graph Learning from Data under Structural and Laplacian Constraints
- Intel RealSense Stereoscopic Depth Cameras