Imperial College London


Faculty of EngineeringDepartment of Computing

Chair in Machine Learning and Pattern Recognition



m.bronstein Website




569Huxley BuildingSouth Kensington Campus






BibTex format

author = {Levie, R and Monti, F and Bresson, X and Bronstein, MM},
doi = {10.1109/TSP.2018.2879624},
journal = {IEEE Transactions on Signal Processing},
pages = {97--109},
title = {CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters},
url = {},
volume = {67},
year = {2018}

RIS format (EndNote, RefMan)

AB - © 1991-2012 IEEE. The rise of graph-structured data such as social networks, regulatory networks, citation graphs, and functional brain networks, in combination with resounding success of deep learning in various applications, has brought the interest in generalizing deep learning models to non-Euclidean domains. In this paper, we introduce a new spectral domain convolutional architecture for deep learning on graphs. The core ingredient of our model is a new class of parametric rational complex functions (Cayley polynomials) allowing to efficiently compute spectral filters on graphs that specialize on frequency bands of interest. Our model generates rich spectral filters that are localized in space, scales linearly with the size of the input data for sparsely connected graphs, and can handle different constructions of Laplacian operators. Extensive experimental results show the superior performance of our approach, in comparison to other spectral domain convolutional architectures, on spectral image classification, community detection, vertex classification, and matrix completion tasks.
AU - Levie,R
AU - Monti,F
AU - Bresson,X
AU - Bronstein,MM
DO - 10.1109/TSP.2018.2879624
EP - 109
PY - 2018///
SN - 1053-587X
SP - 97
TI - CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filters
T2 - IEEE Transactions on Signal Processing
UR -
VL - 67
ER -