Skip to Main content Skip to Navigation
Conference papers

Nested Learning for Multi-Level Classification

Abstract : Deep neural networks models are generally designed and trained for a specific type and quality of data. In this work, we address this problem in the context of nested learning. For many applications, both the input data, at training and testing, and the prediction can be conceived at multiple nested quality/resolutions. We show that by leveraging this multiscale information, the problem of poor generalization and prediction overconfidence, as well as the exploitation of multiple training data quality, can be efficiently addressed. We evaluate the proposed ideas in six public datasets: MNIST, Fashion-MNIST, CIFAR10, CIFAR100, Plantvillage, and DBPEDIA. We observe that coarsely annotated data can help to solve fine predictions and reduce overconfidence significantly. We also show that hierarchical learning produces models intrinsically more robust to adversarial attacks and data perturbations.
Complete list of metadata
Contributor : Raphael Achddou Connect in order to contact the contributor
Submitted on : Tuesday, March 23, 2021 - 9:55:02 AM
Last modification on : Tuesday, October 19, 2021 - 11:15:16 AM


Files produced by the author(s)


  • HAL Id : hal-03177336, version 1



Raphaël Achddou, J. Matias Di Martino, Guillermo Sapiro. Nested Learning for Multi-Level Classification. ICASSP, Jun 2021, Toronto (virtuel), Canada. ⟨hal-03177336⟩



Les métriques sont temporairement indisponibles