Geometric Deep Learning: Going beyond Euclidean data
By Michael M. Bronstein, Joan Bruna, Yann LeCun, Arthur Szlam, Pierre Vandergheynst. 2016
3,894 citations recorded by the source corpus.
Many scientific fields study data with an underlying structure that is non-Euclidean. Some examples include social networks in computational social sciences, sensor networks in communications, functional networks in brain imaging, regulatory networks in genetics, and meshed surfaces in computer graphics. In many applications, such geometric data are large and complex (in the case of social networks, on the scale of billions) and are natural targets for machine-learning techniques. In particular, we would like to use deep neural networks, which have recently proven to be powerful tools for a broad range of problems from computer vision, natural-language processing, and audio analysis. However, these tools have been most successful on data with an underlying Euclidean or grid-like structure and in cases where the invariances of these structures are built into networks used to model them.
Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition, Model Architectures, Graph Learning
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computational Mechanics
- Andreas Geiger
- Michael M. Bronstein
- Leonidas J. Guibas
- Yaron Lipman
- Or Litany
- M. Ovsjanikov
- Michael Niemeyer
- Matthias Nießner
In Computer Graphics and Computer-Aided Design
- Andreas Geiger
- Michael Niemeyer
- Yaron Lipman
- Michael M. Bronstein
- Matthias Nießner
- Lars M. Mescheder
- Yiyi Liao
- M. Ovsjanikov
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Trevor Darrell
- Ming-Hsuan Yang
- Xiaogang Wang
- Jitendra Malik
- Andrea Vedaldi
- Andrew Zisserman
In Model Architectures
- Yoshua Bengio
- Quoc V. Le
- Kyunghyun Cho
- Ruslan Salakhutdinov
- J. Urgen Schmidhuber
- Max Welling
- Mohit Bansal
- Sergey Levine
In Graph Learning
- William L. Hamilton
- Jure Leskovec
- Stephan Günnemann
- Michael M. Bronstein
- Matthias Fey
- Aleksandar Bojchevski
- Rex Ying
- Jian Tang
Citations in the corpus
- Local Relation Networks for Image Recognition (2019, neutral): “Self-Attention / Graph Networks The proposed local relation layer is also related to self-attention models [26] used in natural language processing, and to graph networks applied on non-grid data [3].”
- Perceiver: General Perception with Iterative Attention (2021, background): “, 2019), or for groups other than the group of linear translations (e.g. Cohen & Welling 2016; Bronstein et al. 2017; Esteves et al. 2018)); this however often requires a tremendous amount of researcher time and expertise.”
- Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs (2016, background): “∗Equal contribution keen interest in geometric deep learning [11] attempting to generalize such methods to non-Euclidean structure data.”
- Graph Convolutional Neural Networks for Web-Scale Recommender Systems (2018, methodology): “, to exploit user-to-item interaction graphs as well as social graphs) [6, 19, 21, 24, 29, 30].”
- Fast Graph Representation Learning with PyTorch Geometric (2019, background): “Similar to the concepts of convolutional and pooling layers on regular domains, GNNs are able to (hierarchically) extract localized embeddings by passing, transforming, and aggregating information (Bronstein et al., 2017; Gilmer et al., 2017; Battaglia et al., 2018; Ying et al., 2018).”
- Occupancy Networks: Learning 3D Reconstruction in Function Space (2018, background): “Mesh Representations: Meshes have first been considered for discriminative 3D classification or segmentation tasks by applying convolutions on the graph spanned by the mesh’s vertices and edges [5, 27, 71].”
- RealPoint3D: Point Cloud Generation from a Single Image with Complex Background (2018, methodology): “Beside point clouds, many frameworks have been proposed for meshes representation, such as (Bronstein et al. 2016), spectral graph CNN (Yi et al. 2017) and Geodesic CNN (GCNN) (Masci et al. 2015).”
- On the Binding Problem in Artificial Neural Networks (2020, background): “Graph Convolutional Networks Graph Convolutional Networks (GCNs) are a type of GNNs based on a generalization of convolutional neural networks (which operate on grids) to non-Euclidean geometries such as graphs (Bronstein et al., 2017).”
- Recurrent Independent Mechanisms (2019, methodology): “Relational Graph Based Methods: Recent graph-based architectures have studied combinatorial generalization in the context of modeling dynamical systems like physics simulation, multi-object scenes, and motion-capture data, and multiagent systems (Scarselli et al., 2008; Bronstein et al., 2017; Watte”
- Structured agents for physical construction (2019, background): “, 2014), implemented with graph neural networks (Scarselli et al., 2009; Bronstein et al., 2017; Gilmer et al., 2017; Battaglia et al., 2018).”
- Relational inductive biases, deep learning, and graph networks (2018, background): “Recently, a class of models has arisen at the intersection of deep learning and structured approaches, which focuses on approaches for reasoning about explicitly structured data, in particular graphs (e.g. Scarselli et al., 2009b; Bronstein et al., 2017; Gilmer et al., 2017; Wang et al., 2018c; Li e”
- Flexible Neural Representation for Physics Prediction (2018, positive): “Our work draws inspiration from and extends that of Chang et al. [11] and Battaglia et al. [4], which in turn use ideas from graph-based neural networks [39, 44, 9, 30, 22, 14, 13, 24, 8, 40].”
- Deep Graph Infomax (2018, methodology): “Generalizing neural networks to graph-structured inputs is one of the current major challenges of machine learning (Bronstein et al., 2017; Hamilton et al., 2017b; Battaglia et al., 2018).”
- Hierarchical Graph Representation Learning with Differentiable Pooling (2018, neutral): “[4] outlines connections to spectral graph convolutions.”
- Representation Learning on Graphs: Methods and Applications (2017, methodology): “|V| State-of-the-art approximations to these spectral approaches (e.g., using Chebyshev polynomials) are conceptually similar to Algorithm 1, with some minor variations, and we refer the reader to Bronstein et al. [7] for a thorough discussion of these techniques.”
- Cormorant: Covariant Molecular Neural Networks (2019, background): “Similar ideas underly generalizations of CNNs to manifolds [Masci et al., 2015, Monti et al., 2016, Bronstein et al., 2017] and graphs [Bruna et al., 2014, Henaff et al., 2015].”
- Graphite: Iterative Generative Modeling of Graphs (2018, background): “In contrast, graphs are discrete, combinatorial objects whose underlying structure lies on a non-Euclidean space (Bronstein et al., 2017).”
- GraphVAE: Towards Generation of Small Graphs Using Variational Autoencoders (2018, neutral): “Deep learning on graphs has very recently become a popular research topic (Bronstein et al., 2017), with useful applications across fields such as chemistry (Gilmer et al., 2017), medicine (Ktena et al., 2017), or computer vision (Si-monovsky & Komodakis, 2017).”
- Gauge Equivariant Convolutional Networks and the Icosahedral CNN (2019, background): “Geometric Deep Learning Geometric deep learning (Bronstein et al., 2017) is concerned with the generalization of (convolutional) neural networks to manifolds.”