Learning Image-Based Contaminant Detection in Wool Fleece from Noisy Annotations

Patten, T; Alempijevic, A; Fitch, R

Learning Image-Based Contaminant Detection in Wool Fleece from Noisy Annotations

Patten, T

Alempijevic, A

Fitch, R

Permalink

Publisher:: Springer International Publishing
Publication Type:: Chapter
Citation:: Computer Vision Systems, 2021, 12899 LNCS, pp. 234-244
Issue Date:: 2021-01-01

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download Accepted versionAdobe PDF (5.97 MB)

View on publisher's site

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Patten, T https://orcid.org/0000-0003-1139-9451
dc.contributor.author	Alempijevic, A https://orcid.org/0000-0002-1769-8041
dc.contributor.author	Fitch, R https://orcid.org/0000-0003-2215-5188
dc.date.accessioned	2022-01-13T00:42:07Z
dc.date.available	2022-01-13T00:42:07Z
dc.date.issued	2021-01-01
dc.identifier.citation	Computer Vision Systems, 2021, 12899 LNCS, pp. 234-244
dc.identifier.isbn	9783030871550
dc.identifier.uri	http://hdl.handle.net/10453/153019
dc.description.abstract	This paper addresses the problem of detecting natural contaminants in freshly shorn wool fleece in RGB images using deep learning-based semantic segmentation. The challenge of inconsistent annotation is overcome by learning the probability of contamination as opposed to a discrete class. From the continuous value predictions, contaminated regions can be extracted by selectively thresholding on the probability of contamination. Furthermore, the imbalance of the class distributions is accounted for by adaptively weighting each pixel’s contribution to the loss function. Results show that the adaptive weight improves the prediction accuracy and overall outperforms learning an approximated representation by quantising the distributions.
dc.language	en
dc.publisher	Springer International Publishing
dc.relation.ispartof	Computer Vision Systems
dc.relation.isbasedon	10.1007/978-3-030-87156-7_19
dc.rights	info:eu-repo/semantics/openAccess
dc.subject.classification	Artificial Intelligence & Image Processing
dc.title	Learning Image-Based Contaminant Detection in Wool Fleece from Noisy Annotations
dc.type	Chapter
utslib.citation.volume	12899 LNCS
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Strength - CAS - Centre for Autonomous Systems
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Mechanical and Mechatronic Engineering
utslib.copyright.status	open_access	*
dc.date.updated	2022-01-13T00:42:04Z
pubs.publication-status	Published
pubs.volume	12899 LNCS

Abstract:

This paper addresses the problem of detecting natural contaminants in freshly shorn wool fleece in RGB images using deep learning-based semantic segmentation. The challenge of inconsistent annotation is overcome by learning the probability of contamination as opposed to a discrete class. From the continuous value predictions, contaminated regions can be extracted by selectively thresholding on the probability of contamination. Furthermore, the imbalance of the class distributions is accounted for by adaptively weighting each pixel’s contribution to the loss function. Results show that the adaptive weight improves the prediction accuracy and overall outperforms learning an approximated representation by quantising the distributions.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/153019