Deakin University
Browse

File(s) under permanent embargo

Iterative learning with open-set noisy labels

conference contribution
posted on 2018-01-01, 00:00 authored by Yisen Wang, Weiyang Liu, Daniel Ma, James Bailey, Hongyuan Zha, Le Song, Shu-Tao Xia
Large-scale datasets possessing clean label annotations are crucial for training Convolutional Neural Networks (CNNs). However, labeling large-scale data can be very costly and error-prone, and even high-quality datasets are likely to contain noisy (incorrect) labels. Existing works usually employ a closed-set assumption, whereby the samples associated with noisy labels possess a true class contained within the set of known classes in the training data. However, such an assumption is too restrictive for many applications, since samples associated with noisy labels might in fact possess a true class that is not present in the training data. We refer to this more complex scenario as the open-set noisy label problem and show that it is nontrivial in order to make accurate predictions. To address this problem, we propose a novel iterative learning framework for training CNNs on datasets with open-set noisy labels. Our approach detects noisy labels and learns deep discriminative features in an iterative fashion. To benefit from the noisy label detection, we design a Siamese network to encourage clean labels and noisy labels to be dissimilar. A reweighting module is also applied to simultaneously emphasize the learning from clean labels and reduce the effect caused by noisy labels. Experiments on CIFAR-10, ImageNet and real-world noisy (web-search) datasets demonstrate that our proposed model can robustly train CNNs in the presence of a high proportion of open-set as well as closed-set noisy labels.

History

Event

Computer vision and pattern recognition. Conference (2018 : Salt Lake City, Ut.)

Series

IEEE Computer Society Conference

Pagination

8688 - 8696

Publisher

Institute of Electrical and Electronics Engineers

Location

Salt Lake City, Ut.

Place of publication

Piscataway, N.J.

Start date

2018-06-18

End date

2018-06-23

ISSN

1063-6919

eISSN

2575-7075

ISBN-13

9781538664209

Language

eng

Publication classification

E1.1 Full written paper - refereed

Editor/Contributor(s)

Unknown

Title of proceedings

CVPRW 2018 : Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops