{"id":{"repo_id":"uts","oai_identifier":"oai:opus.lib.uts.edu.au:10453/160984"},"canonical_url":"https://search.dev.ndltd.org/etd/uts/oai:opus.lib.uts.edu.au:10453/160984","repository":{"repo_id":"uts","name":"University of Technology Sydney","base_url":"https://opus.lib.uts.edu.au/oai/request"},"display":{"title":"Automated and Handcrafted Neural Network Design for Vision Applications","abstract":"This thesis presents effective neural network design for various vision applications from two aspects, automated neural network design and handcrafted neural network design. To be more specific, in the first step, it applies novel neural architecture search algorithms on single image deraining and re-identification (reID), where unique deraining and reID search space are proposed, respectively. To step further, this thesis also introduces elaborately handcrafted networks, such as a holistic LSTM with extra designed memory cells and gated operations for pedestrian trajectory prediction and a hybrid transformer-convolutional network for video deraining. To facilitate the applications of designed neural networks, the thesis further presents a semi-supervised learning based framework that leverages only a few labelled data instead of relying on tedious data annotations during training.","abstract_html":"This thesis presents effective neural network design for various vision applications from two aspects, automated neural network design and handcrafted neural network design. To be more specific, in the first step, it applies novel neural architecture search algorithms on single image deraining and re-identification (reID), where unique deraining and reID search space are proposed, respectively. To step further, this thesis also introduces elaborately handcrafted networks, such as a holistic LSTM with extra designed memory cells and gated operations for pedestrian trajectory prediction and a hybrid transformer-convolutional network for video deraining. To facilitate the applications of designed neural networks, the thesis further presents a semi-supervised learning based framework that leverages only a few labelled data instead of relying on tedious data annotations during training.","abstract_has_math":false,"creators":["Quan, Ruijie"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022","date_published":"2022","updated_at":"2026-07-24T06:32:04Z","subjects":[],"languages":["en_US"],"rights":["The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.","au.edu.uts.lib/ppc","info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/10453/160984","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Quan, Ruijie"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2022-08-28T23:17:43Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2022-08-28T23:17:43Z"]},{"key":"dc:date.issued","label":"Date","values":["2022"]},{"key":"dc:relation","label":"Dc Relation","values":["https://opus.lib.uts.edu.au/bitstream/10453/160984/2/02whole.pdf"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en_US"]},{"key":"dc:rights","label":"Dc Rights","values":["The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.","au.edu.uts.lib/ppc","info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://hdl.handle.net/10453/160984"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["University of Technology Sydney. Faculty of Engineering and Information Technology."]},{"key":"dc:description.abstract","label":"Abstract","values":["This thesis presents effective neural network design for various vision applications from two aspects, automated neural network design and handcrafted neural network design. To be more specific, in the first step, it applies novel neural architecture search algorithms on single image deraining and re-identification (reID), where unique deraining and reID search space are proposed, respectively. To step further, this thesis also introduces elaborately handcrafted networks, such as a holistic LSTM with extra designed memory cells and gated operations for pedestrian trajectory prediction and a hybrid transformer-convolutional network for video deraining. To facilitate the applications of designed neural networks, the thesis further presents a semi-supervised learning based framework that leverages only a few labelled data instead of relying on tedious data annotations during training."]},{"key":"dc:format","label":"Dc Format","values":["Thesis (PhD)"]},{"key":"dc:title","label":"Title","values":["Automated and Handcrafted Neural Network Design for Vision Applications"]}]}],"canonical_facts":{"dc:creator":["Quan, Ruijie"],"dc:date.accessioned":["2022-08-28T23:17:43Z"],"dc:date.available":["2022-08-28T23:17:43Z"],"dc:date.issued":["2022"],"dc:description":["University of Technology Sydney. Faculty of Engineering and Information Technology."],"dc:description.abstract":["This thesis presents effective neural network design for various vision applications from two aspects, automated neural network design and handcrafted neural network design. To be more specific, in the first step, it applies novel neural architecture search algorithms on single image deraining and re-identification (reID), where unique deraining and reID search space are proposed, respectively. To step further, this thesis also introduces elaborately handcrafted networks, such as a holistic LSTM with extra designed memory cells and gated operations for pedestrian trajectory prediction and a hybrid transformer-convolutional network for video deraining. To facilitate the applications of designed neural networks, the thesis further presents a semi-supervised learning based framework that leverages only a few labelled data instead of relying on tedious data annotations during training."],"dc:format":["Thesis (PhD)"],"dc:identifier.uri":["http://hdl.handle.net/10453/160984"],"dc:language.iso":["en_US"],"dc:relation":["https://opus.lib.uts.edu.au/bitstream/10453/160984/2/02whole.pdf"],"dc:rights":["The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.","au.edu.uts.lib/ppc","info:eu-repo/semantics/openAccess"],"dc:title":["Automated and Handcrafted Neural Network Design for Vision Applications"],"dc:type":["Thesis"]},"updated_at":"2026-07-24T06:32:04Z"}