Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges, Michael M. Bronstein, Joan Bruna, Taco Cohen, Petar Veličković, 2021 (MIT Press) - A comprehensive and modern textbook that provides a unified theoretical framework for geometric deep learning, with extensive coverage of Graph Signal Processing as a foundation for Graph Neural Networks.
Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering, Michaël Defferrard, Xavier Bresson, Pierre Vandergheynst, 2016Advances in Neural Information Processing Systems 29 (NeurIPS) - Introduces the use of Chebyshev polynomials to efficiently approximate graph convolutional filters, enabling localized and scalable spectral GNNs (ChebNet).