NLNL: Negative Learning for Noisy Labels

By Youngdong Kim, Junho Yim, Juseung Yun, Junmo Kim. IEEE International Conference on Computer Vision, 2019

348 citations recorded by the source corpus.

Convolutional Neural Networks (CNNs) provide excellent performance when used for image classification. The classical method of training CNNs is by labeling images in a supervised manner as in ``input image belongs to this label'' (Positive Learning; PL), which is a fast and accurate method if the labels are assigned correctly to all images. However, if inaccurate labels, or noisy labels, exist, training with PL will provide wrong information, thus severely degrading performance. To address this issue, we start with an indirect learning method called Negative Learning (NL), in which the CNNs are trained using a complementary label as in ``input image does not belong to this complementary label.'' Because the chances of selecting a true label as a complementary label are low, NL decreases the risk of providing incorrect information. Furthermore, to improve convergence, we extend our method by adopting PL selectively, termed as Selective Negative Learning and Positive Learning (SelNLPL). PL is used selectively to train upon expected-to-be-clean data, whose choices become possible as NL progresses, thus resulting in superior performance of filtering out noisy data. With simple semi-supervised training technique, our method achieves state-of-the-art accuracy for noisy data classification, proving the superiority of SelNLPL's noisy data filtering ability.

Subfields: Artificial Intelligence, Civil and Structural Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/ab7539b938ca8a99b5fb34695e1b86ef0c6f3632

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Civil and Structural Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Normalized Loss Functions for Deep Learning with Noisy Labels (2020, methodology): “Inspired by the beneﬁt of symmetric (Wang et al., 2019c) or complementary learning (Kim et al., 2019), we propose to combine a robust active loss and a robust passive loss into an “Active Passive Loss” (APL) framework for both robust and sufﬁcient learning.”
A Comprehensive Survey on Test-Time Adaptation Under Distribution Shifts (2023, neutral): “Motivated by the idea of negative learning [192], PR-SFDA [193] randomly chooses a label from the set {1, .”
Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization (2020, background): “In addition to the aforementioned approaches, there are some other deep learning solutions [13, 17] to deal with noisy labels, including pseudo-label based [35, 40] and robust loss based approaches [28, 46].”
Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey (2019, methodology): “oost [126], noise detection based AdaBoost [127], rBoost [128], RBoost1&RBoost2 [129], robust multi-class AdaBoost [130] 7.Others Minimum covariance determinant [131], complementary labels [132], [133], dataset with noisy and less noisy labels [134], autoencoder reconstruction error [135] TABLE ”
In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning (2021, methodology): “On the other hand, Kim et al. (2019) use negative learning primarily to obtain good network initializations to learn with noisy labels.”
Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels (2022, background): “Negative Learning aims at decreasing the risk of incorrect information by lowering the probability of negative samples [26,27,35,40], but those negative samples are selected with high confidence.”