Deep learning models of biological visual information processing
Improved computational models of biological vision can shed light on key processes contributing to the high accuracy of the human visual system. Deep learning models, which extract multiple layers of increasingly complex features from data, achieved recent breakthroughs on visual tasks. This thesis...
| Main Author: | |
|---|---|
| Format: | Thesis (University of Nottingham only) |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | https://eprints.nottingham.ac.uk/35561/ |
| _version_ | 1848795107919134720 |
|---|---|
| author | Turcsány, Diána |
| author_facet | Turcsány, Diána |
| author_sort | Turcsány, Diána |
| building | Nottingham Research Data Repository |
| collection | Online Access |
| description | Improved computational models of biological vision can shed light on key processes contributing to the high accuracy of the human visual system. Deep learning models, which extract multiple layers of increasingly complex features from data, achieved recent breakthroughs on visual tasks. This thesis proposes such flexible data-driven models of biological vision and also shows how insights regarding biological visual processing can lead to advances within deep learning.
To harness the potential of deep learning for modelling the retina and early vision, this work introduces a new dataset and a task simulating an early visual processing function and evaluates deep belief networks (DBNs) and deep neural networks (DNNs) on this input. The models are shown to learn feature detectors similar to retinal ganglion and V1 simple cells and execute early vision tasks.
To model high-level visual information processing, this thesis proposes novel deep learning architectures and training methods. Biologically inspired Gaussian receptive field constraints are imposed on restricted Boltzmann machines (RBMs) to improve the fidelity of the data representation to encodings extracted by visual processing neurons. Moreover, concurrently with learning local features, the proposed local receptive field constrained RBMs (LRF-RBMs) automatically discover advantageous non-uniform feature detector placements from data.
Following the hierarchical organisation of the visual cortex, novel LRF-DBN and LRF-DNN models are constructed using LRF-RBMs with gradually increasing receptive field sizes to extract consecutive layers of features. On a challenging face dataset, unlike DBNs, LRF-DBNs learn a feature hierarchy exhibiting hierarchical part-based composition. Also, the proposed deep models outperform DBNs and DNNs on face completion and dimensionality reduction, thereby demonstrating the strength of methods inspired by biological visual processing. |
| first_indexed | 2025-11-14T19:26:50Z |
| format | Thesis (University of Nottingham only) |
| id | nottingham-35561 |
| institution | University of Nottingham Malaysia Campus |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-14T19:26:50Z |
| publishDate | 2016 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | nottingham-355612025-02-28T11:50:15Z https://eprints.nottingham.ac.uk/35561/ Deep learning models of biological visual information processing Turcsány, Diána Improved computational models of biological vision can shed light on key processes contributing to the high accuracy of the human visual system. Deep learning models, which extract multiple layers of increasingly complex features from data, achieved recent breakthroughs on visual tasks. This thesis proposes such flexible data-driven models of biological vision and also shows how insights regarding biological visual processing can lead to advances within deep learning. To harness the potential of deep learning for modelling the retina and early vision, this work introduces a new dataset and a task simulating an early visual processing function and evaluates deep belief networks (DBNs) and deep neural networks (DNNs) on this input. The models are shown to learn feature detectors similar to retinal ganglion and V1 simple cells and execute early vision tasks. To model high-level visual information processing, this thesis proposes novel deep learning architectures and training methods. Biologically inspired Gaussian receptive field constraints are imposed on restricted Boltzmann machines (RBMs) to improve the fidelity of the data representation to encodings extracted by visual processing neurons. Moreover, concurrently with learning local features, the proposed local receptive field constrained RBMs (LRF-RBMs) automatically discover advantageous non-uniform feature detector placements from data. Following the hierarchical organisation of the visual cortex, novel LRF-DBN and LRF-DNN models are constructed using LRF-RBMs with gradually increasing receptive field sizes to extract consecutive layers of features. On a challenging face dataset, unlike DBNs, LRF-DBNs learn a feature hierarchy exhibiting hierarchical part-based composition. Also, the proposed deep models outperform DBNs and DNNs on face completion and dimensionality reduction, thereby demonstrating the strength of methods inspired by biological visual processing. 2016-10-15 Thesis (University of Nottingham only) NonPeerReviewed application/pdf en arr https://eprints.nottingham.ac.uk/35561/1/thesis_DianaTurcsany.pdf Turcsány, Diána (2016) Deep learning models of biological visual information processing. PhD thesis, University of Nottingham. deep learning machine learning visual information processing biological vision retinal modelling neural computation local receptive field constrained restricted Boltzmann machine deep neural network deep belief network deep autoencoder feature hub self-adaptive structure structure learning face completion |
| spellingShingle | deep learning machine learning visual information processing biological vision retinal modelling neural computation local receptive field constrained restricted Boltzmann machine deep neural network deep belief network deep autoencoder feature hub self-adaptive structure structure learning face completion Turcsány, Diána Deep learning models of biological visual information processing |
| title | Deep learning models of biological visual information processing |
| title_full | Deep learning models of biological visual information processing |
| title_fullStr | Deep learning models of biological visual information processing |
| title_full_unstemmed | Deep learning models of biological visual information processing |
| title_short | Deep learning models of biological visual information processing |
| title_sort | deep learning models of biological visual information processing |
| topic | deep learning machine learning visual information processing biological vision retinal modelling neural computation local receptive field constrained restricted Boltzmann machine deep neural network deep belief network deep autoencoder feature hub self-adaptive structure structure learning face completion |
| url | https://eprints.nottingham.ac.uk/35561/ |