{"id":{"repo_id":"uwo","oai_identifier":"oai:uwo.scholaris.ca:20.500.14721/39622"},"canonical_url":"https://search.dev.ndltd.org/etd/uwo/oai:uwo.scholaris.ca:20.500.14721/39622","repository":{"repo_id":"uwo","name":"Western University","base_url":"https://uwo.scholaris.ca/server/oai/request"},"display":{"title":"Improving Controllability in Diffusion-Based Image Inpainting through Structured Workflows and Preference-Based Model Adaptation","abstract":"This thesis presents two complementary approaches for improving controllability in diffusion-based image generation. The first addresses limitations of existing inpainting-based virtual staging methods, which can produce objects that are semantically inappropriate, visually unconvincing, or poorly matched to the surrounding environment. To address this, an end-to-end prompt-based contextual virtual staging framework is developed that fine-tunes a diffusion model for controlled object generation and integrates it into an agentic pipeline for context-aware scene inpainting. Building on this, the second approach addresses the limitation that workflow-level control alone does not ensure that the underlying inpainting model fully internalizes prompt and scene context. To address this, a pretrained diffusion inpainting model is directly fine-tuned using preference-based optimization, together with a composite reward model and a region-aware evaluation system. Together, these approaches improve controllability at both the workflow level and the model level.","abstract_html":"This thesis presents two complementary approaches for improving controllability in diffusion-based image generation. The first addresses limitations of existing inpainting-based virtual staging methods, which can produce objects that are semantically inappropriate, visually unconvincing, or poorly matched to the surrounding environment. To address this, an end-to-end prompt-based contextual virtual staging framework is developed that fine-tunes a diffusion model for controlled object generation and integrates it into an agentic pipeline for context-aware scene inpainting. Building on this, the second approach addresses the limitation that workflow-level control alone does not ensure that the underlying inpainting model fully internalizes prompt and scene context. To address this, a pretrained diffusion inpainting model is directly fine-tuned using preference-based optimization, together with a composite reward model and a region-aware evaluation system. Together, these approaches improve controllability at both the workflow level and the model level.","abstract_has_math":false,"creators":["Murray, Scott Jason"],"institution":"The University of Western Ontario","degree_name":"M Eng Sci","degree_level":null,"degree_discipline":"Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":[],"advisors":["Yang, Yimin"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026-04-14","date_published":"2026-04-14","updated_at":"2026-07-27T21:56:16Z","subjects":["Diffusion Models","Image Inpainting","Controllable Image Editing","Preference Optimization","Reinforcement Learning","Virtual Staging","Generative AI","Stable Diffusion","LoRA Fine-Tuning","Context-Aware Generation"],"languages":["en"],"rights":["Attribution-NonCommercial-ShareAlike 4.0 International"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/20.500.14721/39622","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Yang, Yimin"]},{"key":"dc:creator","label":"Author","values":["Murray, Scott Jason"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-04-30T19:19:40Z"]},{"key":"dc:date.issued","label":"Date","values":["2026-04-14"]},{"key":"dc:publisher","label":"Institution","values":["The University of Western Ontario"]},{"key":"dc:type","label":"Dc Type","values":["thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical and Computer Engineering"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M Eng Sci"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["The University of Western Ontario"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Diffusion Models","Image Inpainting","Controllable Image Editing","Preference Optimization","Reinforcement Learning","Virtual Staging","Generative AI","Stable Diffusion","LoRA Fine-Tuning","Context-Aware Generation"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Attribution-NonCommercial-ShareAlike 4.0 International"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/20.500.14721/39622"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["This thesis presents two complementary approaches for improving controllability in diffusion-based image generation. The first addresses limitations of existing inpainting-based virtual staging methods, which can produce objects that are semantically inappropriate, visually unconvincing, or poorly matched to the surrounding environment. To address this, an end-to-end prompt-based contextual virtual staging framework is developed that fine-tunes a diffusion model for controlled object generation and integrates it into an agentic pipeline for context-aware scene inpainting. Building on this, the second approach addresses the limitation that workflow-level control alone does not ensure that the underlying inpainting model fully internalizes prompt and scene context. To address this, a pretrained diffusion inpainting model is directly fine-tuned using preference-based optimization, together with a composite reward model and a region-aware evaluation system. Together, these approaches improve controllability at both the workflow level and the model level."]},{"key":"dc:title","label":"Title","values":["Improving Controllability in Diffusion-Based Image Inpainting through Structured Workflows and Preference-Based Model Adaptation"]}]}],"canonical_facts":{"dc:contributor.advisor":["Yang, Yimin"],"dc:creator":["Murray, Scott Jason"],"dc:date.accessioned":["2026-04-30T19:19:40Z"],"dc:date.issued":["2026-04-14"],"dc:description.abstract":["This thesis presents two complementary approaches for improving controllability in diffusion-based image generation. The first addresses limitations of existing inpainting-based virtual staging methods, which can produce objects that are semantically inappropriate, visually unconvincing, or poorly matched to the surrounding environment. To address this, an end-to-end prompt-based contextual virtual staging framework is developed that fine-tunes a diffusion model for controlled object generation and integrates it into an agentic pipeline for context-aware scene inpainting. Building on this, the second approach addresses the limitation that workflow-level control alone does not ensure that the underlying inpainting model fully internalizes prompt and scene context. To address this, a pretrained diffusion inpainting model is directly fine-tuned using preference-based optimization, together with a composite reward model and a region-aware evaluation system. Together, these approaches improve controllability at both the workflow level and the model level."],"dc:identifier.uri":["https://hdl.handle.net/20.500.14721/39622"],"dc:language.iso":["en"],"dc:publisher":["The University of Western Ontario"],"dc:rights":["Attribution-NonCommercial-ShareAlike 4.0 International"],"dc:subject":["Diffusion Models","Image Inpainting","Controllable Image Editing","Preference Optimization","Reinforcement Learning","Virtual Staging","Generative AI","Stable Diffusion","LoRA Fine-Tuning","Context-Aware Generation"],"dc:title":["Improving Controllability in Diffusion-Based Image Inpainting through Structured Workflows and Preference-Based Model Adaptation"],"dc:type":["thesis"],"thesis:degree_discipline":["Electrical and Computer Engineering"],"thesis:degree_name":["M Eng Sci"],"thesis:institution_name":["The University of Western Ontario"]},"updated_at":"2026-07-27T21:56:16Z"}