Scale-invariant learning and convolutional networks

DOI10.1016/J.ACHA.2016.06.005MaRDI QIDQ347531zbMATH OpenOpenAlexFDO

Authors Soumith Chintala, Marc'Aurelio Ranzato, Yuandong Tian, Mark Tygert, Wojciech Zaremba, Arthur Szlam

Publication date 30 November 2016

Published in Applied and Computational Harmonic Analysis (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/j.acha.2016.06.005

zbMATH Keywords

waveletsclassificationmachine learningsignal processingimage processingspectrumwavelet packetsequivariantinvariantrepresentation

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Learning and adaptive systems in artificial intelligence (68T05) Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Numerical methods for wavelets (65T60) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08)

Recommendations

• Scale-covariant and scale-invariant Gaussian derivative networks
• Scale equivariant neural networks with morphological scale-spaces
• Unsupervised learning of invariant representations
• Scale-invariant unconstrained online learning
• Provably scale-covariant continuous hierarchical networks based on scale-normalized differential expressions coupled in cascade

Cites work

• A framework for kernel-based multi-category classification
• A mathematical motivation for complex-valued convolutional networks
• Multicategory vertex discriminant analysis for high-dimensional data
• Nonlinear vertex discriminant analysis with reproducing kernels

Cited in

(10)

• Simple convolutional-based models: are they learning the task or the data?
• A mathematical motivation for complex-valued convolutional networks
• Scale-covariant and scale-invariant Gaussian derivative networks
• Scale-covariant and scale-invariant Gaussian derivative networks
• Can generalised divergences help for invariant neural networks?
• Scale-invariant scale-channel networks: deep networks that generalise to previously unseen scales
• Provably scale-covariant continuous hierarchical networks based on scale-normalized differential expressions coupled in cascade
• Scale equivariant neural networks with morphological scale-spaces
• Why do deep convolutional networks generalize so poorly to small image transformations?
• A model of cardinality blindness in inferotemporal cortex

Describes a project that uses

Uses Software

• CIFAR
• ElemStatLearn

This page was built for publication: Scale-invariant learning and convolutional networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q347531)