{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/124347"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/124347","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Exploring knowledge in generative models","abstract":"Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms","abstract_html":"Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms","abstract_has_math":false,"creators":["Bhattad, Anand"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Forsyth, David A","Efros, Alexei A","Hoiem, Derek W","Wang, Shenlong","Lazebnik, Svetlana","Freeman, William T"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024-04-20","date_published":"2024-04-20","updated_at":"2026-07-22T22:25:00Z","subjects":["Visual Knowledge","Generative Models","Intrinsic Images","Relighting","Image Decomposition","Normals","Depth","Albedo","Shading","Segmentation"],"languages":["en","eng"],"rights":["Copyright 2024 Anand Bhattad"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2142/124347","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Forsyth, David A","Efros, Alexei A","Hoiem, Derek W","Wang, Shenlong","Lazebnik, Svetlana","Freeman, William T"]},{"key":"dc:creator","label":"Author","values":["Bhattad, Anand"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2024-04-20","2024-05"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Visual Knowledge","Generative Models","Intrinsic Images","Relighting","Image Decomposition","Normals","Depth","Albedo","Shading","Segmentation"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en","eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2024 Anand Bhattad"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/2142/124347"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms","The student, Anand Bhattad, accepted the attached license on 2024-04-19 at 13:10.","The student, Anand Bhattad, submitted this Dissertation for approval on 2024-04-19 at 13:19.","This Dissertation was approved for publication on 2024-04-20 at 10:35.","DSpace SAF Submission Ingestion Package generated from Vireo submission #20502 on 2024-09-16 at 00:35:30","Generative models, such as StyleGAN, have demonstrated remarkable ability in producing realistic and controllable images. However, the underlying representations and mechanisms employed by these models remain largely unexplored. This thesis delves into the intrinsic properties and manipulability of StyleGAN, focusing on image relighting and decomposition. We begin by exploring the impact of image decompositions on image-based relighting. By analyzing the role of intrinsic image components such as reflectance, shading, and normals, we gain insights into the fundamental properties that contribute to realistic relighting. This understanding lays the foundation for our subsequent investigations into StyleGAN. Building upon these insights, we introduce StyLitGAN, a method that enables StyleGAN to generate scenes with novel lighting conditions. StyLitGAN produces realistic lighting effects, including cast shadows, soft shadows, inter-reflections, and glossy effects, without requiring labeled, paired, or CGI data. Moreover, it seamlessly extends to manipulating surface properties like colors and materials. Next, we present Make It So, a near-perfect GAN inversion technique that significantly outperforms previous state-of-the-art methods. Make It So can invert and relight real scenes, including out-of-domain images, demonstrating its generalizability and robustness. Finally, we uncover hidden gems within StyleGAN, providing strong evidence that it encodes easily accessible and accurate internal representations of familiar scene properties, known as ``intrinsic images,\" as defined by Barrow and Tenenbaum in their seminal work from 1978. We demonstrate that StyleGAN has encodings for intrinsic images such as reflectance, shading, and normals, which can be extracted and manipulated for various applications. Through our discoveries, we shed light on the implicit understanding of worldly knowledge present within generative models like StyleGAN. Our findings pave the way for improved manipulability, understanding, and refinement of generative models, with potential applications in computer vision, computational photography, computer graphics, and machine learning. This thesis contributes to the broader goal of leveraging generative models for advanced image manipulation and scene understanding tasks."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Exploring knowledge in generative models"]}]}],"canonical_facts":{"dc:contributor":["Forsyth, David A","Efros, Alexei A","Hoiem, Derek W","Wang, Shenlong","Lazebnik, Svetlana","Freeman, William T"],"dc:creator":["Bhattad, Anand"],"dc:date":["2024-04-20","2024-05"],"dc:description":["Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms","The student, Anand Bhattad, accepted the attached license on 2024-04-19 at 13:10.","The student, Anand Bhattad, submitted this Dissertation for approval on 2024-04-19 at 13:19.","This Dissertation was approved for publication on 2024-04-20 at 10:35.","DSpace SAF Submission Ingestion Package generated from Vireo submission #20502 on 2024-09-16 at 00:35:30","Generative models, such as StyleGAN, have demonstrated remarkable ability in producing realistic and controllable images. However, the underlying representations and mechanisms employed by these models remain largely unexplored. This thesis delves into the intrinsic properties and manipulability of StyleGAN, focusing on image relighting and decomposition. We begin by exploring the impact of image decompositions on image-based relighting. By analyzing the role of intrinsic image components such as reflectance, shading, and normals, we gain insights into the fundamental properties that contribute to realistic relighting. This understanding lays the foundation for our subsequent investigations into StyleGAN. Building upon these insights, we introduce StyLitGAN, a method that enables StyleGAN to generate scenes with novel lighting conditions. StyLitGAN produces realistic lighting effects, including cast shadows, soft shadows, inter-reflections, and glossy effects, without requiring labeled, paired, or CGI data. Moreover, it seamlessly extends to manipulating surface properties like colors and materials. Next, we present Make It So, a near-perfect GAN inversion technique that significantly outperforms previous state-of-the-art methods. Make It So can invert and relight real scenes, including out-of-domain images, demonstrating its generalizability and robustness. Finally, we uncover hidden gems within StyleGAN, providing strong evidence that it encodes easily accessible and accurate internal representations of familiar scene properties, known as ``intrinsic images,\" as defined by Barrow and Tenenbaum in their seminal work from 1978. We demonstrate that StyleGAN has encodings for intrinsic images such as reflectance, shading, and normals, which can be extracted and manipulated for various applications. Through our discoveries, we shed light on the implicit understanding of worldly knowledge present within generative models like StyleGAN. Our findings pave the way for improved manipulability, understanding, and refinement of generative models, with potential applications in computer vision, computational photography, computer graphics, and machine learning. This thesis contributes to the broader goal of leveraging generative models for advanced image manipulation and scene understanding tasks."],"dc:format":["application/pdf"],"dc:identifier":["https://hdl.handle.net/2142/124347"],"dc:language":["en","eng"],"dc:rights":["Copyright 2024 Anand Bhattad"],"dc:subject":["Visual Knowledge","Generative Models","Intrinsic Images","Relighting","Image Decomposition","Normals","Depth","Albedo","Shading","Segmentation"],"dc:title":["Exploring knowledge in generative models"],"dc:type":["text"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:00Z"}