Skip to content

A Framework Using Contrastive Learning for Classification with Noisy Labels

We propose a framework using contrastive learning as a pre-training task to perform image classification in the presence of noisy labels. Recent strategies, such as pseudo-labelling, sample selection with Gaussian Mixture models, and weighted supervised contrastive learning have been combined into a fine-tuning phase following the pre-training. In this paper, we provide an extensive empirical study showing that a preliminary contrastive learning step brings a significant gain in performance when using different loss functions: non robust, robust, and early-learning regularized. Our experiments performed on standard benchmarks and real-world datasets demonstrate that: (i) the contrastive pre-training increases the robustness of any loss function to noisy labels and (ii) the additional fine-tuning phase can further improve accuracy but at the cost of additional complexity.

Madalina Ciortan, Romain Dupuis and Thomas Peel, A Framework Using Contrastive Learning for Classification with Noisy Labels, Data, 2021, 6, 61.

DOI: https://doi.org/10.3390/data6060061

Watch the presentation on YouTube.

Public implementation: https://github.com/ciortanmadalina/constrastive-noisy-label

Click here to access the paper.

Releated Posts

We Collaborate on the TAUDoS Project

We started a new collaboration with Aix-Marseille University, Montreal University, Nantes University, and St-Etienne on a four-year project called TAUDoS, which focuses on Trustful AI.
Read More

DEBS 2022

In June 2022, our research director Sabri Skhiri and the head of the data science department at Madalina Ciortan travelled to Copenhagen to attend DEBS 2022, the leading conference focusing on distributed and event-based systems.
Read More