Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
The class imbalance problem in computer vision
Thesis (MSc)--Stellenbosch University, 2022.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | en_ZA |
| Published: |
Stellenbosch : Stellenbosch University
2022
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613975857856512 |
|---|---|
| access_status_str | Open Access |
| author | Crous, Willem Hendrik |
| author2 | Brink, Willie |
| author_browse | Brink, Willie Crous, Willem Hendrik |
| author_facet | Brink, Willie Crous, Willem Hendrik |
| author_sort | Crous, Willem Hendrik |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description | Thesis (MSc)--Stellenbosch University, 2022. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/124719 |
| institution | Stellenbosch University (South Africa) |
| language | en_ZA |
| last_indexed | 2026-06-10T12:44:41.678Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2022 |
| publishDateRange | 2022 |
| publishDateSort | 2022 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/124719 The class imbalance problem in computer vision Crous, Willem Hendrik Brink, Willie Stellenbosch University. Faculty of Science. Dept. of Mathematical Sciences (Applied Mathematics) Computer vision Image classification Class imbalance UCTD Data sets -- Characteristic Thesis (MSc)--Stellenbosch University, 2022. ENGLISH ABSTRACT: Class imbalance is a naturally occurring phenomenon, typically characterised as a dataset consisting of classes with varying numbers of samples. When trained on class imbalanced data, networks tend to favour frequently occurring (majority) classes over the less frequent (minority) classes. This poses chal- lenges for tasks reliant upon accurate recognition of the less frequent classes. The aim of this thesis is to investigate general methods towards addressing this problem. First we establish why a network may favour majority classes. We contend that as less frequent classes are likely to under-represent the re- quired underlying distribution for a given task, training may produce a decision boundary that transgresses the feature space of minority classes. Additionally we find that the weight norms of the classification layer in a neural network may tend towards the distribution of the training data, thus affecting the de- cision boundary. We determine that this decision boundary shift impacts both the accuracy and confidence calibration of neural networks. We investigate several approaches to shift the decision boundary. The first approach acquires additional data and increases the representation of minority classes. This is achieved through either creating synthetic samples following a distribution- aware regularisation method, or utilising additional unlabelled data in a semi- supervised setting. The second approach aims to adjust the classifier weight norms by separately training the classifier and feature extractor. We find that implementing an effective regularisation method with a simple decoupled sam- pling scheme can provide considerable improvements over standard sampling methods. Furthermore we find that utilising additional unlabelled data may lead to additional gains given certain dataset characteristics are taken into consideration. AFRIKAANSE OPSOMMING: Geen opsomming beskikbaar Masters 2022-02-23T08:03:54Z 2022-04-29T09:28:18Z 2022-02-23T08:03:54Z 2022-04-29T09:28:18Z 2022-04 Thesis http://hdl.handle.net/10019.1/124719 en_ZA Stellenbosch University 76 pages application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Computer vision Image classification Class imbalance UCTD Data sets -- Characteristic Crous, Willem Hendrik The class imbalance problem in computer vision |
| title | The class imbalance problem in computer vision |
| title_full | The class imbalance problem in computer vision |
| title_fullStr | The class imbalance problem in computer vision |
| title_full_unstemmed | The class imbalance problem in computer vision |
| title_short | The class imbalance problem in computer vision |
| title_sort | class imbalance problem in computer vision |
| topic | Computer vision Image classification Class imbalance UCTD Data sets -- Characteristic |
| url | http://hdl.handle.net/10019.1/124719 |
| work_keys_str_mv | AT crouswillemhendrik theclassimbalanceproblemincomputervision AT crouswillemhendrik classimbalanceproblemincomputervision |