Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Frequency-domain deconvolution in deep learning
This dissertation presents an exhaustive exploration of a novel approach to deep learning in computer vision tasks: the frequency-domain deconvolution operation. Recognizing the unparalleled success of convolutional neural networks (CNNs) in the realm of computer vision, we critically evaluate the p...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English English |
| Published: |
Department of Electrical Engineering
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613312961740800 |
|---|---|
| access_status_str | Open Access |
| author | Meyer, Moegamad |
| author2 | Nicolls, Frederick |
| author_browse | Meyer, Moegamad Nicolls, Frederick |
| author_facet | Nicolls, Frederick Meyer, Moegamad |
| author_sort | Meyer, Moegamad |
| collection | Thesis |
| description | This dissertation presents an exhaustive exploration of a novel approach to deep learning in computer vision tasks: the frequency-domain deconvolution operation. Recognizing the unparalleled success of convolutional neural networks (CNNs) in the realm of computer vision, we critically evaluate the performance and computational demands of traditional convolution operations against the proposed deconvolution method. Using a systematic approach, we apply the deconvolution layer to two quintessential computer vision problems: image classification and single image super resolution (SISR). The results demonstrate the deconvolution layer's potential in certain scenarios, with marked improvements observed in image classification. For SISR tasks, though advantages were noticed under specific configurations, the traditional CNNs still demonstrated their robustness. Additionally, the dissertation touches upon the layer's computational demands, revealing an increased computational overhead for the deconvolution layer. Encouragingly, the layer demonstrated promising attributes like learning long-range filters and isolating objects from backgrounds effectively. Concluding with avenues for future research, this dissertation acts as a stepping stone in the uncharted territory of deconvolution operations, emphasising innovation alongside evaluation in the dynamic world of deep learning. |
| format | Thesis |
| id | oai:open.uct.ac.za:11427/41745 |
| institution | University of Cape Town (South Africa) |
| language | English eng |
| last_indexed | 2026-06-10T12:34:08.683Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UCTD — University of Cape Town Open Access Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Department of Electrical Engineering |
| publisherStr | Department of Electrical Engineering |
| record_format | dspace |
| source_str | UCTD — University of Cape Town Open Access Repository |
| spelling | oai:open.uct.ac.za:11427/41745 Frequency-domain deconvolution in deep learning Meyer, Moegamad Nicolls, Frederick Convolutional neural networks This dissertation presents an exhaustive exploration of a novel approach to deep learning in computer vision tasks: the frequency-domain deconvolution operation. Recognizing the unparalleled success of convolutional neural networks (CNNs) in the realm of computer vision, we critically evaluate the performance and computational demands of traditional convolution operations against the proposed deconvolution method. Using a systematic approach, we apply the deconvolution layer to two quintessential computer vision problems: image classification and single image super resolution (SISR). The results demonstrate the deconvolution layer's potential in certain scenarios, with marked improvements observed in image classification. For SISR tasks, though advantages were noticed under specific configurations, the traditional CNNs still demonstrated their robustness. Additionally, the dissertation touches upon the layer's computational demands, revealing an increased computational overhead for the deconvolution layer. Encouragingly, the layer demonstrated promising attributes like learning long-range filters and isolating objects from backgrounds effectively. Concluding with avenues for future research, this dissertation acts as a stepping stone in the uncharted territory of deconvolution operations, emphasising innovation alongside evaluation in the dynamic world of deep learning. 2025-09-10T10:46:42Z 2025-09-10T10:46:42Z 2025 2025-09-10T10:22:26Z Thesis / Dissertation Masters MSc http://hdl.handle.net/11427/41745 en eng application/pdf Department of Electrical Engineering Faculty of Engineering and the Built Environment University of Cape Town |
| spellingShingle | Convolutional neural networks Meyer, Moegamad Frequency-domain deconvolution in deep learning |
| thesis_degree_str | Master's |
| title | Frequency-domain deconvolution in deep learning |
| title_full | Frequency-domain deconvolution in deep learning |
| title_fullStr | Frequency-domain deconvolution in deep learning |
| title_full_unstemmed | Frequency-domain deconvolution in deep learning |
| title_short | Frequency-domain deconvolution in deep learning |
| title_sort | frequency domain deconvolution in deep learning |
| topic | Convolutional neural networks |
| url | http://hdl.handle.net/11427/41745 |
| work_keys_str_mv | AT meyermoegamad frequencydomaindeconvolutionindeeplearning |