Graph-based Isometry Invariant Representation Learning

By Renata Khasanova, Pascal Frossard. 2017

55 citations recorded by the source corpus.

Learning transformation invariant representations of visual data is an important problem in computer vision. Deep convolutional networks have demonstrated remarkable results for image and video classification tasks. However, they have achieved only limited success in the classification of images that undergo geometric transformations. In this work we present a novel Transformation Invariant Graph-based Network (TIGraNet), which learns graph-based features that are inherently invariant to isometric transformations such as rotation and translation of input images. In particular, images are represented as signals on graphs, which permits to replace classical convolution and pooling layers in deep networks with graph spectral convolution and dynamic graph pooling layers that together contribute to invariance to isometric transformations. Our experiments show high performance on rotated and translated images from the test set compared to classical architectures that are very sensitive to transformations in the data. The inherent invariance properties of our framework provide key advantages, such as increased resiliency to data variability and sustained performance with limited training sets.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Conditional variance penalties and domain shift robustness (2017, background): “invariances from data and the focus often lies on geometric transformations of the input data or explicitly creating augmented observations (Sohn and Lee, 2012; Khasanova and Frossard, 2017; Hashimoto et al., 2017; Devries and Taylor, 2017).”
SphereNet: Learning Spherical Representations for Detection and Classification in Omnidirectional Images (2018, background): “While a graph representation solves the problem of discontinuities at the borders of an equirectangular image, graph convolutional networks are limited to small graphs and image resolutions (50 × 50 pixels in [15]) and have not yet demonstrated recognition performance comparable to regular CNNs on m”
Edge Attention-based Multi-Relational Graph Convolutional Networks (2018, methodology): “Since edge attentions are shared across all graphs, our EAGCN method also extracts invariant properties of graphs [20].”
Graph Signal Processing: Overview, Challenges, and Applications (2017, methodology): “Here, the GSP framework leads to the development of fast localized convolutional filters on graphs [209] along with adapted pooling operators [210].”
Graph-Based Classification of Omnidirectional Images (2017, methodology): “Luckily graph-based deep learning techniques have been recently introduced [13, 14, 15] that allow applying deep learning techniques to irregularly structured data.”
Guarding Against Adversarial Domain Shifts with Counterfactual Regularization (2017, background): “ input.jpg output.jpg to the image. 33 invariances from data and the focus often lies on geometric transformations of the input data or explicitly creating augmented observations (Sohn and Lee, 2012; Khasanova and Frossard, 2017; Hashimoto et al., 2017; Devries and Taylor, 2017). The main dierence b”
Graph Wavelet Neural Network (2019, background): “Some subsequent works make spectral methods spectrum-free (Defferrard et al., 2016; Kipf & Welling, 2017; Khasanova & Frossard, 2017), achieving locality in spatial domain and avoiding high computational cost of the eigendecomposition of Laplacian matrix.”