Human Interfaces with Machine Learning Recognition Systems

CLARKSON, CONNOR

doi:10.23889/SUThesis.69942

E-Thesis 292 views 349 downloads

Human Interfaces with Machine Learning Recognition Systems / CONNOR CLARKSON

Swansea University Author: CONNOR CLARKSON

PDF | E-Thesis – open access

Copyright: The author, Connor Clarkson, 2025 Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0).
Download (6.95MB)

DOI (Published version): 10.23889/SUThesis.69942

Abstract

Building large pools of data has become a relatively straightforward task, with many automated ways of obtaining different sources of data. Labelling such data has resulted in becoming an exponential problem, both in terms of time and in the form of an interaction-heavy task. This task only becomes...

Full description

Published:	Swansea University, Wales, UK 2025
Institution:	Swansea University
Degree level:	Doctoral
Degree name:	Ph.D
Supervisor:	Xianghua, X., and Edwards, M.
URI:	https://cronfa.swan.ac.uk/Record/cronfa69942

first_indexed	2025-07-10T13:02:00Z
last_indexed	2025-07-11T05:02:56Z
id	cronfa69942
recordtype	RisThesis
fullrecord	<?xml version="1.0"?><rfc1807><datestamp>2025-07-10T16:05:52.9655972</datestamp><bib-version>v2</bib-version><id>69942</id><entry>2025-07-10</entry><title>Human Interfaces with Machine Learning Recognition Systems</title><swanseaauthors><author><sid>09de10d9b005815dc6341cd4a062eaf4</sid><firstname>CONNOR</firstname><surname>CLARKSON</surname><name>CONNOR CLARKSON</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2025-07-10</date><abstract>Building large pools of data has become a relatively straightforward task, with many automated ways of obtaining different sources of data. Labelling such data has resulted in becoming an exponential problem, both in terms of time and in the form of an interaction-heavy task. This task only becomes exponential with feature-rich structures of data and labelling systems, as well as requiring more advanced expertise for many different domains of a task to model. A prominent set of techniques utilising this data, and large networks have reformed machine learning into what we call deep learning today. Within this ﬁeld, we can form levels of supervision that allow for stronger signals of inductive bias for both deep network architectures and in the training scheme. In this work, we explore both types with the target application and domain being the manufacturing of steel. Firstly, we present an exploratory approach to assist in decision-making for the task of clustering by utilising the feature-rich representations provided by generative models. By forming it as a semi-supervised problem we can provide varying degrees of supervision to enhance performance as a form of inductive bias into the training scheme. Supervision can be formalised into labels from data or in an active learning setting where we request help from an expert. If we are required to make a request, then we must provide information and visualisations so that an accurate decision can be made. Following this, in our second body of work we extend on an active learning setting by introducing a new acquisition function based on the distance from different representations. We apply it to a data reﬁnement strategy where we ﬁx mistakes in bounding-box labelled datasets to form a dense segmentation. Different forms of user interaction provide different levels of information to the training scheme, we explore the effects of these user interactions on the performance of this reﬁnement task. Lastly, we apply stronger forms of inductive bias into the network architecture by modelling hierarchical labelling systems, where such relationships between labels form an abstraction and ﬁne-grained level of the data. Inspired by the structure of human cognition and perception where we recognise patterns of various levels of abstraction to deﬁne an object. By invoking an explicit form of deep learning with feature-rich structures like graphs we can model these interconnected labels. We deﬁne two types of hierarchical relationships: the ﬁrst is a break-up of the physical or geometric structure of the object, referred to as an encapsulation relationship. The second is sub-classiﬁcation relationships which are semantic relations of labels provided by domain knowledge of what we are trying to capture in the dataset. We utilise both to solve classiﬁcation and segmentation tasks.</abstract><type>E-Thesis</type><journal/><volume/><journalNumber/><paginationStart/><paginationEnd/><publisher/><placeOfPublication>Swansea University, Wales, UK</placeOfPublication><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic/><keywords/><publishedDay>6</publishedDay><publishedMonth>5</publishedMonth><publishedYear>2025</publishedYear><publishedDate>2025-05-06</publishedDate><doi>10.23889/SUThesis.69942</doi><url/><notes>A selection of content is redacted or is partially redacted from this thesis to protect sensitive and personal information.</notes><college>COLLEGE NANME</college><CollegeCode>COLLEGE CODE</CollegeCode><institution>Swansea University</institution><supervisor>Xianghua, X., and Edwards, M.</supervisor><degreelevel>Doctoral</degreelevel><degreename>Ph.D</degreename><degreesponsorsfunders>EPSRC Centre For Doctoral Training in Enhancing Human Interactions and Collaborations with Data and Intelligence Driven Systems</degreesponsorsfunders><apcterm/><funders>EPSRC Centre For Doctoral Training in Enhancing Human Interactions and Collaborations with Data and Intelligence Driven Systems</funders><projectreference/><lastEdited>2025-07-10T16:05:52.9655972</lastEdited><Created>2025-07-10T13:52:27.9587037</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>CONNOR</firstname><surname>CLARKSON</surname><order>1</order></author></authors><documents><document><filename>69942__34740__cd07d56c71c54a6e8a9a2f0ea4bb47ad.pdf</filename><originalFilename>2025_Clarkson_C.final.69942.pdf</originalFilename><uploaded>2025-07-10T16:05:19.8587463</uploaded><type>Output</type><contentLength>7285872</contentLength><contentType>application/pdf</contentType><version>E-Thesis – open access</version><cronfaStatus>true</cronfaStatus><documentNotes>Copyright: The author, Connor Clarkson, 2025 Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0).</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling	2025-07-10T16:05:52.9655972 v2 69942 2025-07-10 Human Interfaces with Machine Learning Recognition Systems 09de10d9b005815dc6341cd4a062eaf4 CONNOR CLARKSON CONNOR CLARKSON true false 2025-07-10 Building large pools of data has become a relatively straightforward task, with many automated ways of obtaining different sources of data. Labelling such data has resulted in becoming an exponential problem, both in terms of time and in the form of an interaction-heavy task. This task only becomes exponential with feature-rich structures of data and labelling systems, as well as requiring more advanced expertise for many different domains of a task to model. A prominent set of techniques utilising this data, and large networks have reformed machine learning into what we call deep learning today. Within this ﬁeld, we can form levels of supervision that allow for stronger signals of inductive bias for both deep network architectures and in the training scheme. In this work, we explore both types with the target application and domain being the manufacturing of steel. Firstly, we present an exploratory approach to assist in decision-making for the task of clustering by utilising the feature-rich representations provided by generative models. By forming it as a semi-supervised problem we can provide varying degrees of supervision to enhance performance as a form of inductive bias into the training scheme. Supervision can be formalised into labels from data or in an active learning setting where we request help from an expert. If we are required to make a request, then we must provide information and visualisations so that an accurate decision can be made. Following this, in our second body of work we extend on an active learning setting by introducing a new acquisition function based on the distance from different representations. We apply it to a data reﬁnement strategy where we ﬁx mistakes in bounding-box labelled datasets to form a dense segmentation. Different forms of user interaction provide different levels of information to the training scheme, we explore the effects of these user interactions on the performance of this reﬁnement task. Lastly, we apply stronger forms of inductive bias into the network architecture by modelling hierarchical labelling systems, where such relationships between labels form an abstraction and ﬁne-grained level of the data. Inspired by the structure of human cognition and perception where we recognise patterns of various levels of abstraction to deﬁne an object. By invoking an explicit form of deep learning with feature-rich structures like graphs we can model these interconnected labels. We deﬁne two types of hierarchical relationships: the ﬁrst is a break-up of the physical or geometric structure of the object, referred to as an encapsulation relationship. The second is sub-classiﬁcation relationships which are semantic relations of labels provided by domain knowledge of what we are trying to capture in the dataset. We utilise both to solve classiﬁcation and segmentation tasks. E-Thesis Swansea University, Wales, UK 6 5 2025 2025-05-06 10.23889/SUThesis.69942 A selection of content is redacted or is partially redacted from this thesis to protect sensitive and personal information. COLLEGE NANME COLLEGE CODE Swansea University Xianghua, X., and Edwards, M. Doctoral Ph.D EPSRC Centre For Doctoral Training in Enhancing Human Interactions and Collaborations with Data and Intelligence Driven Systems EPSRC Centre For Doctoral Training in Enhancing Human Interactions and Collaborations with Data and Intelligence Driven Systems 2025-07-10T16:05:52.9655972 2025-07-10T13:52:27.9587037 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science CONNOR CLARKSON 1 69942__34740__cd07d56c71c54a6e8a9a2f0ea4bb47ad.pdf 2025_Clarkson_C.final.69942.pdf 2025-07-10T16:05:19.8587463 Output 7285872 application/pdf E-Thesis – open access true Copyright: The author, Connor Clarkson, 2025 Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0). true eng https://creativecommons.org/licenses/by/4.0/
title	Human Interfaces with Machine Learning Recognition Systems
spellingShingle	Human Interfaces with Machine Learning Recognition Systems CONNOR CLARKSON
title_short	Human Interfaces with Machine Learning Recognition Systems
title_full	Human Interfaces with Machine Learning Recognition Systems
title_fullStr	Human Interfaces with Machine Learning Recognition Systems
title_full_unstemmed	Human Interfaces with Machine Learning Recognition Systems
title_sort	Human Interfaces with Machine Learning Recognition Systems
author_id_str_mv	09de10d9b005815dc6341cd4a062eaf4
author_id_fullname_str_mv	09de10d9b005815dc6341cd4a062eaf4_***_CONNOR CLARKSON
author	CONNOR CLARKSON
author2	CONNOR CLARKSON
format	E-Thesis
publishDate	2025
institution	Swansea University
doi_str_mv	10.23889/SUThesis.69942
college_str	Faculty of Science and Engineering
hierarchytype
hierarchy_top_id	facultyofscienceandengineering
hierarchy_top_title	Faculty of Science and Engineering
hierarchy_parent_id	facultyofscienceandengineering
hierarchy_parent_title	Faculty of Science and Engineering
department_str	School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
document_store_str	1
active_str	0
description	Building large pools of data has become a relatively straightforward task, with many automated ways of obtaining different sources of data. Labelling such data has resulted in becoming an exponential problem, both in terms of time and in the form of an interaction-heavy task. This task only becomes exponential with feature-rich structures of data and labelling systems, as well as requiring more advanced expertise for many different domains of a task to model. A prominent set of techniques utilising this data, and large networks have reformed machine learning into what we call deep learning today. Within this ﬁeld, we can form levels of supervision that allow for stronger signals of inductive bias for both deep network architectures and in the training scheme. In this work, we explore both types with the target application and domain being the manufacturing of steel. Firstly, we present an exploratory approach to assist in decision-making for the task of clustering by utilising the feature-rich representations provided by generative models. By forming it as a semi-supervised problem we can provide varying degrees of supervision to enhance performance as a form of inductive bias into the training scheme. Supervision can be formalised into labels from data or in an active learning setting where we request help from an expert. If we are required to make a request, then we must provide information and visualisations so that an accurate decision can be made. Following this, in our second body of work we extend on an active learning setting by introducing a new acquisition function based on the distance from different representations. We apply it to a data reﬁnement strategy where we ﬁx mistakes in bounding-box labelled datasets to form a dense segmentation. Different forms of user interaction provide different levels of information to the training scheme, we explore the effects of these user interactions on the performance of this reﬁnement task. Lastly, we apply stronger forms of inductive bias into the network architecture by modelling hierarchical labelling systems, where such relationships between labels form an abstraction and ﬁne-grained level of the data. Inspired by the structure of human cognition and perception where we recognise patterns of various levels of abstraction to deﬁne an object. By invoking an explicit form of deep learning with feature-rich structures like graphs we can model these interconnected labels. We deﬁne two types of hierarchical relationships: the ﬁrst is a break-up of the physical or geometric structure of the object, referred to as an encapsulation relationship. The second is sub-classiﬁcation relationships which are semantic relations of labels provided by domain knowledge of what we are trying to capture in the dataset. We utilise both to solve classiﬁcation and segmentation tasks.
published_date	2025-05-06T05:29:32Z
_version_	1851097950072078336
score	11.089386

Human Interfaces with Machine Learning Recognition Systems / CONNOR CLARKSON

Similar Items