Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge
Noisy data, crawled from the web or supplied by volunteers such as Mechanical Turkers or citizen scientists, is considered an alternative to professionally labeled data. There has been research focused on mitigating the effects of label noise. It is typically modeled as inaccuracy, where the correct...
Main Authors: | , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
arXiv
2020
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2010.06469 https://arxiv.org/abs/2010.06469 |
id |
ftdatacite:10.48550/arxiv.2010.06469 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2010.06469 2023-05-15T18:20:04+02:00 Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge Brust, Clemens-Alexander Barz, Björn Denzler, Joachim 2020 https://dx.doi.org/10.48550/arxiv.2010.06469 https://arxiv.org/abs/2010.06469 unknown arXiv arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Computer Vision and Pattern Recognition cs.CV Machine Learning cs.LG FOS Computer and information sciences Article CreativeWork article Preprint 2020 ftdatacite https://doi.org/10.48550/arxiv.2010.06469 2022-03-10T15:05:57Z Noisy data, crawled from the web or supplied by volunteers such as Mechanical Turkers or citizen scientists, is considered an alternative to professionally labeled data. There has been research focused on mitigating the effects of label noise. It is typically modeled as inaccuracy, where the correct label is replaced by an incorrect label from the same set. We consider an additional dimension of label noise: imprecision. For example, a non-breeding snow bunting is labeled as a bird. This label is correct, but not as precise as the task requires. Standard softmax classifiers cannot learn from such a weak label because they consider all classes mutually exclusive, which non-breeding snow bunting and bird are not. We propose CHILLAX (Class Hierarchies for Imprecise Label Learning and Annotation eXtrapolation), a method based on hierarchical classification, to fully utilize labels of any precision. Experiments on noisy variants of NABirds and ILSVRC2012 show that our method outperforms strong baselines by as much as 16.4 percentage points, and the current state of the art by up to 3.9 percentage points. : 9 pages pre-print. Accepted for publication at ICPR 2020 Article in Journal/Newspaper Snow Bunting DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Computer Vision and Pattern Recognition cs.CV Machine Learning cs.LG FOS Computer and information sciences |
spellingShingle |
Computer Vision and Pattern Recognition cs.CV Machine Learning cs.LG FOS Computer and information sciences Brust, Clemens-Alexander Barz, Björn Denzler, Joachim Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
topic_facet |
Computer Vision and Pattern Recognition cs.CV Machine Learning cs.LG FOS Computer and information sciences |
description |
Noisy data, crawled from the web or supplied by volunteers such as Mechanical Turkers or citizen scientists, is considered an alternative to professionally labeled data. There has been research focused on mitigating the effects of label noise. It is typically modeled as inaccuracy, where the correct label is replaced by an incorrect label from the same set. We consider an additional dimension of label noise: imprecision. For example, a non-breeding snow bunting is labeled as a bird. This label is correct, but not as precise as the task requires. Standard softmax classifiers cannot learn from such a weak label because they consider all classes mutually exclusive, which non-breeding snow bunting and bird are not. We propose CHILLAX (Class Hierarchies for Imprecise Label Learning and Annotation eXtrapolation), a method based on hierarchical classification, to fully utilize labels of any precision. Experiments on noisy variants of NABirds and ILSVRC2012 show that our method outperforms strong baselines by as much as 16.4 percentage points, and the current state of the art by up to 3.9 percentage points. : 9 pages pre-print. Accepted for publication at ICPR 2020 |
format |
Article in Journal/Newspaper |
author |
Brust, Clemens-Alexander Barz, Björn Denzler, Joachim |
author_facet |
Brust, Clemens-Alexander Barz, Björn Denzler, Joachim |
author_sort |
Brust, Clemens-Alexander |
title |
Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
title_short |
Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
title_full |
Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
title_fullStr |
Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
title_full_unstemmed |
Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge |
title_sort |
making every label count: handling semantic imprecision by integrating domain knowledge |
publisher |
arXiv |
publishDate |
2020 |
url |
https://dx.doi.org/10.48550/arxiv.2010.06469 https://arxiv.org/abs/2010.06469 |
genre |
Snow Bunting |
genre_facet |
Snow Bunting |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2010.06469 |
_version_ |
1766197540533829632 |