Separating common from salient patterns with Contrastive Representation Learning - Equipe Image, Modélisation, Analyse, GEométrie, Synthèse
Conference Papers Year : 2024

Separating common from salient patterns with Contrastive Representation Learning

Abstract

Contrastive Analysis is a sub-field of Representation Learning that aims at separating common factors of variation between two datasets, a background (i.e., healthy subjects) and a target (i.e., diseased subjects), from the salient factors of variation, only present in the target dataset. Despite their relevance, current models based on Variational Auto-Encoders have shown poor performance in learning semantically-expressive representations. On the other hand, Contrastive Representation Learning has shown tremendous performance leaps in various applications (classification, clustering, etc.). In this work, we propose to leverage the ability of Contrastive Learning to learn semantically expressive representations well adapted for Contrastive Analysis. We reformulate it under the lens of the InfoMax Principle and identify two Mutual Information terms to maximize and one to minimize. We decompose the first two terms into an Alignment and a Uniformity term, as commonly done in Contrastive Learning. Then, we motivate a novel Mutual Information minimization strategy to prevent information leakage between common and salient distributions. We validate our method, called SepCLR, on three visual datasets and three medical datasets, specifically conceived to assess the pattern separation capability in Contrastive Analysis.
Fichier principal
Vignette du fichier
Final.pdf (3.94 Mo) Télécharger le fichier
Origin Files produced by the author(s)
licence
Copyright

Dates and versions

hal-04537203 , version 1 (08-04-2024)

Licence

Copyright

Identifiers

  • HAL Id : hal-04537203 , version 1

Cite

Robin Louiset, Edouard Duchesnay, Antoine Grigis, Pietro Gori. Separating common from salient patterns with Contrastive Representation Learning. International Conference on Learning Representations (ICLR), May 2024, Vienna, Austria. ⟨hal-04537203⟩
356 View
26 Download

Share

More