CVAIMar 13, 2023

Twin Contrastive Learning with Noisy Labels

arXiv:2303.06930v1108 citationsh-index: 36Has Code
Originality Incremental advance
AI Analysis

This addresses the problem of model performance degradation due to noisy labels in classification tasks, representing an incremental advance in robust learning methods.

The paper tackles learning from noisy labels by proposing TCL, a twin contrastive learning model that uses Gaussian mixture models and cross-supervision to detect and handle label errors, achieving a 7.5% improvement on CIFAR-10 with 90% noisy labels.

Learning from noisy data is a challenging task that significantly degenerates the model performance. In this paper, we present TCL, a novel twin contrastive learning model to learn robust representations and handle noisy labels for classification. Specifically, we construct a Gaussian mixture model (GMM) over the representations by injecting the supervised model predictions into GMM to link label-free latent variables in GMM with label-noisy annotations. Then, TCL detects the examples with wrong labels as the out-of-distribution examples by another two-component GMM, taking into account the data distribution. We further propose a cross-supervision with an entropy regularization loss that bootstraps the true targets from model predictions to handle the noisy labels. As a result, TCL can learn discriminative representations aligned with estimated labels through mixup and contrastive learning. Extensive experimental results on several standard benchmarks and real-world datasets demonstrate the superior performance of TCL. In particular, TCL achieves 7.5\% improvements on CIFAR-10 with 90\% noisy label -- an extremely noisy scenario. The source code is available at \url{https://github.com/Hzzone/TCL}.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes