{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/98435"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/98435","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Learning viewer-centered projections for 3D shape completion","abstract":"\"The goal of this study is to determine the effectiveness of different 3D shape representations in learning to generate volumetric shapes using deep neural networks. We propose to automatically reconstruct a 3D model from a single-view image of an object by synthesizing multiple depth images and inferring the volume through multi-view 3D reconstruction. The final output is a 3D mesh inferred without seeing voxels in the training process. This is similar to the intuition that humans remember (and inherently reproduce) 3D shapes without ever \"\"seeing through\"\" the underlying volume – we think of objects as seen from certain viewpoints and 3D structure is a derived concept. Most previous studies have focused on directly learning the voxel representations, deforming exemplars, or utilizing user interaction. In this paper, we want to learn category-independent object shape representations by simultaneously predicting multiple incomplete surfaces in relation to the viewer with the complete 3D structure in mind. Instead of predicting voxels which typically need to be in low resolution, we hypothesize that learning a representation that can consistently produce partial surfaces in a multi-task learning model enables inter-category 3D shape transfer. We perform shape completion in novel categories and evaluate quantitatively using voxel I/U and surface distance metrics. We also report that the learned representation improves 3D shape classification.\"","abstract_html":""The goal of this study is to determine the effectiveness of different 3D shape representations in learning to generate volumetric shapes using deep neural networks. We propose to automatically reconstruct a 3D model from a single-view image of an object by synthesizing multiple depth images and inferring the volume through multi-view 3D reconstruction. The final output is a 3D mesh inferred without seeing voxels in the training process. This is similar to the intuition that humans remember (and inherently reproduce) 3D shapes without ever ""seeing through"" the underlying volume – we think of objects as seen from certain viewpoints and 3D structure is a derived concept. Most previous studies have focused on directly learning the voxel representations, deforming exemplars, or utilizing user interaction. In this paper, we want to learn category-independent object shape representations by simultaneously predicting multiple incomplete surfaces in relation to the viewer with the complete 3D structure in mind. Instead of predicting voxels which typically need to be in low resolution, we hypothesize that learning a representation that can consistently produce partial surfaces in a multi-task learning model enables inter-category 3D shape transfer. We perform shape completion in novel categories and evaluate quantitatively using voxel I/U and surface distance metrics. We also report that the learned representation improves 3D shape classification."","abstract_has_math":false,"creators":["Shin, Daeyun"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Hoiem, Derek"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2017,"date_issued":"2017-09-29T17:57:12Z","date_published":"2017-09-29T17:57:12Z","updated_at":"2026-07-22T22:24:35Z","subjects":["3D shape learning","Deep encoder-decoder networks","Vision for graphics","Multi-view reconstruction"],"languages":["en"],"rights":["Copyright 2017 Daeyun Shin"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/98435","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Hoiem, Derek"]},{"key":"dc:creator","label":"Author","values":["Shin, Daeyun"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2017-09-29T17:57:12Z","2017-07-20","2017-08"]},{"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":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"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":["3D shape learning","Deep encoder-decoder networks","Vision for graphics","Multi-view reconstruction"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2017 Daeyun Shin"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/98435"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["\"The goal of this study is to determine the effectiveness of different 3D shape representations in learning to generate volumetric shapes using deep neural networks. We propose to automatically reconstruct a 3D model from a single-view image of an object by synthesizing multiple depth images and inferring the volume through multi-view 3D reconstruction. The final output is a 3D mesh inferred without seeing voxels in the training process. This is similar to the intuition that humans remember (and inherently reproduce) 3D shapes without ever \"\"seeing through\"\" the underlying volume – we think of objects as seen from certain viewpoints and 3D structure is a derived concept. Most previous studies have focused on directly learning the voxel representations, deforming exemplars, or utilizing user interaction. In this paper, we want to learn category-independent object shape representations by simultaneously predicting multiple incomplete surfaces in relation to the viewer with the complete 3D structure in mind. Instead of predicting voxels which typically need to be in low resolution, we hypothesize that learning a representation that can consistently produce partial surfaces in a multi-task learning model enables inter-category 3D shape transfer. We perform shape completion in novel categories and evaluate quantitatively using voxel I/U and surface distance metrics. We also report that the learned representation improves 3D shape classification.\"","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2017-09-29 without embargo terms","The student, Daeyun Shin, accepted the attached license on 2017-07-20 at 11:48.","The student, Daeyun Shin, submitted this Thesis for approval on 2017-07-20 at 11:52.","This Thesis was approved for publication on 2017-07-20 at 13:18.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11551 on 2017-09-29 at 11:32:25","Made available in DSpace on 2017-09-29T17:57:12Z (GMT). No. of bitstreams: 2 SHIN-THESIS-2017.pdf: 1030300 bytes, checksum: 17c0e67fe600c2c6f46af835589b5d1c (MD5) LICENSE.txt: 4208 bytes, checksum: 34d51ba1eb0e3db01fc097e4a3fec3d4 (MD5) Previous issue date: 2017-07-20"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Learning viewer-centered projections for 3D shape completion"]}]}],"canonical_facts":{"dc:contributor":["Hoiem, Derek"],"dc:creator":["Shin, Daeyun"],"dc:date":["2017-09-29T17:57:12Z","2017-07-20","2017-08"],"dc:description":["\"The goal of this study is to determine the effectiveness of different 3D shape representations in learning to generate volumetric shapes using deep neural networks. We propose to automatically reconstruct a 3D model from a single-view image of an object by synthesizing multiple depth images and inferring the volume through multi-view 3D reconstruction. The final output is a 3D mesh inferred without seeing voxels in the training process. This is similar to the intuition that humans remember (and inherently reproduce) 3D shapes without ever \"\"seeing through\"\" the underlying volume – we think of objects as seen from certain viewpoints and 3D structure is a derived concept. Most previous studies have focused on directly learning the voxel representations, deforming exemplars, or utilizing user interaction. In this paper, we want to learn category-independent object shape representations by simultaneously predicting multiple incomplete surfaces in relation to the viewer with the complete 3D structure in mind. Instead of predicting voxels which typically need to be in low resolution, we hypothesize that learning a representation that can consistently produce partial surfaces in a multi-task learning model enables inter-category 3D shape transfer. We perform shape completion in novel categories and evaluate quantitatively using voxel I/U and surface distance metrics. We also report that the learned representation improves 3D shape classification.\"","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2017-09-29 without embargo terms","The student, Daeyun Shin, accepted the attached license on 2017-07-20 at 11:48.","The student, Daeyun Shin, submitted this Thesis for approval on 2017-07-20 at 11:52.","This Thesis was approved for publication on 2017-07-20 at 13:18.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11551 on 2017-09-29 at 11:32:25","Made available in DSpace on 2017-09-29T17:57:12Z (GMT). No. of bitstreams: 2 SHIN-THESIS-2017.pdf: 1030300 bytes, checksum: 17c0e67fe600c2c6f46af835589b5d1c (MD5) LICENSE.txt: 4208 bytes, checksum: 34d51ba1eb0e3db01fc097e4a3fec3d4 (MD5) Previous issue date: 2017-07-20"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/98435"],"dc:language":["en"],"dc:rights":["Copyright 2017 Daeyun Shin"],"dc:subject":["3D shape learning","Deep encoder-decoder networks","Vision for graphics","Multi-view reconstruction"],"dc:title":["Learning viewer-centered projections for 3D shape completion"],"dc:type":["text"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:24:35Z"}