Review–A Survey of Learning from Noisy Labels
Authors:
Open Access
- 1 June 2022
- journal article
- review article
- Published by The Electrochemical Society in ECS Sensors Plus
- Vol. 1 (2) , 021401
Abstract
Deep Learning has achieved remarkable successes in many industry applications and scientific research fields. One essential reason is that deep models can learn rich information from large-scale training datasets through supervised learning. It has been well accepted that the robust deep models heavily rely on the quality of data labels. However, current large-scale datasets mostly involve noisy labels, which are caused by sensor errors, human mistakes, or inaccuracy of search engines, and may severely degrade the performance of deep models. In this survey, we summaries existing works on noisy label learning into two main categories, Loss Correction and Sample Selection, and present their methodologies, commonly used experimental setups, datasets, and the state-of-the-art results. Finally, we discuss a promising research direction that might be valuable for the future study.Keywords
This publication has 15 references indexed in Scilit:
- Image classification with deep learning in the presence of noisy labels: A surveyKnowledge-Based Systems, 2021
- Understanding deep learning (still) requires rethinking generalizationCommunications of the ACM, 2021
- Distilling Effective Supervision From Severe Label Noise2020
- A Novel Self-Supervised Re-labeling Approach for Training with Noisy Labels2020
- Symmetric Cross Entropy for Robust Learning With Noisy Labels2019
- ImageNet classification with deep convolutional neural networksCommunications of the ACM, 2017
- Librispeech: An ASR corpus based on public domain audio books2015
- Noise Tolerance Under Risk MinimizationIEEE Transactions on Cybernetics, 2013
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- On Information and SufficiencyThe Annals of Mathematical Statistics, 1951