Show simple item record

dc.contributor.authorBäuerle, Alexen_US
dc.contributor.authorNeumann, Heikoen_US
dc.contributor.authorRopinski, Timoen_US
dc.contributor.editorViola, Ivan and Gleicher, Michael and Landesberger von Antburg, Tatianaen_US
dc.description.abstractTraining data plays an essential role in modern applications of machine learning. However, gathering labeled training data is time-consuming. Therefore, labeling is often outsourced to less experienced users, or completely automated. This can introduce errors, which compromise valuable training data, and lead to suboptimal training results. We thus propose a novel approach that uses the power of pretrained classifiers to visually guide users to noisy labels, and let them interactively check error candidates, to iteratively improve the training data set. To systematically investigate training data, we propose a categorization of labeling errors into three different types, based on an analysis of potential pitfalls in label acquisition processes. For each of these types, we present approaches to detect, reason about, and resolve error candidates, as we propose measures and visual guidance techniques to support machine learning users. Our approach has been used to spot errors in well-known machine learning benchmark data sets, and we tested its usability during a user evaluation. While initially developed for images, the techniques presented in this paper are independent of the classification algorithm, and can also be extended to many other types of training data.en_US
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.rightsAttribution 4.0 International License
dc.subjectInformation systems
dc.subjectExpert systems
dc.subjectHuman centered computing
dc.subjectUser centered design
dc.subjectInformation visualization
dc.titleClassifier-Guided Visual Correction of Noisy Labels for Image Classification Tasksen_US
dc.description.seriesinformationComputer Graphics Forum
dc.description.sectionheadersMachine Learning

Files in this item


This item appears in the following Collection(s)

  • 39-Issue 3
    EuroVis 2020 - Conference Proceedings

Show simple item record

Attribution 4.0 International License
Except where otherwise noted, this item's license is described as Attribution 4.0 International License