Invariant Scattering Convolution Networks

By Joan Bruna, Stéphane Mallat. 2013

377 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

A Hybrid Differential Evolution Approach to Designing Deep Convolutional Neural Networks for Image Classification (2018, methodology): “on the aforementioned benchmark datasets in the literature [3]. Therefore, they are picked as the peer competitors of the proposed algorithm, which are CAE-2 [13], TIRBM [15], PGBM+DN1 [5], ScatNet-2 [2], RandNet-2 [3], PCANet-2 (softmax) [3], LDANet-2 [3], SVM+RBF [8], SVM+Poly [8], NNet [8], SAA-3”
Building a Regular Decision Boundary with Deep Networks (2017, background): “Deep networks have been shown to be covariant to such actions [1, 16], to linearize them [24, 22, 30] and combing those strategies permit building invariants to them [3, 20, 9].”
Deep Scattering Spectrum with deep neural networks (2014, background): “DSNs take a rawsignal and generate a contractive representation, which preserves signal energy, while ensuring Lipschitz continuity to deformations ( [11]and [12]).”
From BoW to CNN: Two Decades of Texture Representation for Texture Classification (2018, methodology): “2017, 2016b) and ScatNet (Bruna and Mallat 2013), that have rotation and gray scale invariances, give perfect accuracies, revealing one of the limitations of CNN based descriptors in being sensitive to image degradations.”
A survey on face detection in the wild: Past, present and future (2015, background): “Intuitively, following the rational of invariant scattering networks [149], we can postulate that all the layers contribute to the development of features invariant to face deformations.”
Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering (2016, methodology): “The pre-processing is done by a recently introduced technique, namely, the scattering network (ScatNet) (Bruna & Mallat, 2013).”
A Group-Theoretic Framework for Data Augmentation (2019, background): “Related work by Mallat, Bölcskei and others (e.g., Mallat, 2012; Bruna and Mallat, 2013; Wiatowski and Bölcskei, 2018; Anselmi et al., 2019) tries to explain how CNNs extract features, using ideas from harmonic analysis.”
Under review as a conference paper at ICLR 2020 completely free-form sampling in image space : Spatial Transformers (2019, background): “However, this usually requires designs specific to certain kinds of deformation, such as shift, rotation, reflection and scaling (Sifre & Mallat, 2013; Bruna & Mallat, 2013; Kanazawa et al., 2016; Cohen & Welling, 2016; Worrall et al., 2017; Esteves et al., 2018).”
Universality of Deep Convolutional Neural Networks (2018, background): “A main concern about deep learning which has attracted much scientific attention and some criticism is its lack of theories supporting its practical efficiency caused by its network structures, though there have been some theoretical attempts from approximation theory viewpoints [3,19,21,26].”
Interpretable Convolutional Neural Networks via Feedforward Design (2018, background): “Recent studies on CNNs include: visualization of ﬁlter responses at various layers [6, 7, 8], scattering networks [9, 10, 11], tensor analysis [12], generative modeling [13], relevance propagation [14], Taylor decomposition [15], multi-layer convolutional sparse modeling [16] and over-parameterized…”
Machine learning for data-driven discovery in solid Earth geoscience (2019, background): “A scattering transform (76, 77) can provide natural insights in CNNs relevant to geoscience.”