SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernels
Preprint
Abstract
We present Spline-based Convolutional Neural Networks (SplineCNNs), a variant of deep neural networks for irregular structured and geometric input, e.g., graphs or meshes. Our main contribution is a novel convolution operator based on B-splines, that makes the computation time independent from the kernel size due to the local support property of the B-spline basis functions. As a result, we obtain a generalization of the traditional CNN convolution operator by using continuous kernel functions parametrized by a fixed number of trainable weights. In contrast to related approaches that filter in the spectral domain, the proposed method aggregates features purely in the spatial domain. As a main advantage, SplineCNN allows entire end-to-end training of deep architectures, using only the geometric structure as input, instead of handcrafted feature descriptors. For validation, we apply our method on tasks from the fields of image graph classification, shape correspondence and graph node classification, and show that it outperforms or pars state-of-the-art approaches while being significantly faster and having favorable properties like domain-independence.Keywords
All Related Versions
This publication has 6 references indexed in Scilit:
- CayleyNets: Graph Convolutional Neural Networks With Complex Rational Spectral FiltersIEEE Transactions on Signal Processing, 2018
- Deep Functional Maps: Structured Prediction for Dense Shape Correspondence2017
- Geometric Deep Learning: Going beyond Euclidean dataIEEE Signal Processing Magazine, 2017
- SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentation2017
- The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domainsIEEE Signal Processing Magazine, 2013
- Weighted Graph Cuts without Eigenvectors A Multilevel ApproachIEEE Transactions on Pattern Analysis and Machine Intelligence, 2007