Deep Learning from Noisy Labels with Some Adjustments of a Recent Method

Fazekas, István and Fórián, László and Barta, Attila (2023) Deep Learning from Noisy Labels with Some Adjustments of a Recent Method. INFOCOMMUNICATIONS JOURNAL, 15 (SI). pp. 9-12. ISSN 2061-2079

Preview

Text InfocomJournal_2023_SpecISS_ICAI_2.pdf Download (490kB) | Preview

Abstract

In this paper we have used JoCoR, a fairly recent method for learning with label noise, that makes use of two neural networks with a joint loss function using an additional contrastive loss to increase the agreement between them. This method can be extended to more than two networks in a straightforward way. We have carried out experiments on the CIFAR-10 and CIFAR-100 datasets (contaminated by synthetic label noise) with this kind of extension using several contrastive losses. We have concluded that it makes a significant improvement if we use a third network, especially when we use Kullback-Leibler terms for all possible pairs of softmax outputs. Further extension also means some kind of improvement, but in the case of the CIFAR datasets, those were not so significant, maybe except the cases with lower ratio of label noise.

Actions (login required)

Edit Item