{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/99515"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/99515","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Image and video object selection","abstract":"\"Image and video object selection present fundamental research problems in the computer vision field and have many practical applications. They are important technologies in image and video editing, film production, robotics and autonomous driving etc. Previous methods have serious limitations for those tasks for several reasons. First, most of them use some low-level, handcrafted features which are not optimal. Second, they also lack the high-level understanding of \"\"objectness\"\" and semantics. Last but not the least, their generalization ability on unconstrained scenarios is very poor. Recently, deep learning has become the dominant method for computer vision tasks including recognition and detection since it cannot only learn good feature representation in an end-to-end manner but it is also effective at capturing the high-level semantics. However, its exploration in image and video object selection is still impoverished. Therefore, in this thesis we propose several novel deep-learning based methods to tackle the limitations in image and video object selection. Our algorithms are easy to understand and effective. Experimental results clearly demonstrate the superiority of our algorithms over previous methods. Some highlights include the following: (1) Our interactive segmentation algorithm is the first deep-learning based algorithm and achieves the state-of-the-art results on both small-scale and large-scale benchmarks. (2) Our rectangle-based algorithm novelly transforms rectangle inputs to attention-like distance maps and achieves robust performance for sloppy user selections or misplaced detection boxes. (3) Our image matting algorithm is the first to demonstrate the feasibility of learning an alpha matte end-to-end given an image and trimap. It also achieves state-of-the-art results on image matting and video matting benchmarks. (4) Our video object segmentation method combines CNN network with RNN memory cells to learn both good image feature representation and the temporal-spatial coherence.\"","abstract_html":""Image and video object selection present fundamental research problems in the computer vision field and have many practical applications. They are important technologies in image and video editing, film production, robotics and autonomous driving etc. Previous methods have serious limitations for those tasks for several reasons. First, most of them use some low-level, handcrafted features which are not optimal. Second, they also lack the high-level understanding of ""objectness"" and semantics. Last but not the least, their generalization ability on unconstrained scenarios is very poor. Recently, deep learning has become the dominant method for computer vision tasks including recognition and detection since it cannot only learn good feature representation in an end-to-end manner but it is also effective at capturing the high-level semantics. However, its exploration in image and video object selection is still impoverished. Therefore, in this thesis we propose several novel deep-learning based methods to tackle the limitations in image and video object selection. Our algorithms are easy to understand and effective. Experimental results clearly demonstrate the superiority of our algorithms over previous methods. Some highlights include the following: (1) Our interactive segmentation algorithm is the first deep-learning based algorithm and achieves the state-of-the-art results on both small-scale and large-scale benchmarks. (2) Our rectangle-based algorithm novelly transforms rectangle inputs to attention-like distance maps and achieves robust performance for sloppy user selections or misplaced detection boxes. (3) Our image matting algorithm is the first to demonstrate the feasibility of learning an alpha matte end-to-end given an image and trimap. It also achieves state-of-the-art results on image matting and video matting benchmarks. (4) Our video object segmentation method combines CNN network with RNN memory cells to learn both good image feature representation and the temporal-spatial coherence."","abstract_has_math":false,"creators":["Xu, Ning"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Electrical & Computer Engr","degree_department":null,"school":null,"contributors":["Huang, Thomas","Hasegawa-Johnson, Mark","Lazebnik, Svetlana","Liang, Zhi-Pei"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2018,"date_issued":"2018-03-13T17:35:49Z","date_published":"2018-03-13T17:35:49Z","updated_at":"2026-07-22T22:24:37Z","subjects":["Video segmentation","Object selection","Computer vision","Deep learning","Image segmentation"],"languages":["en"],"rights":["Copyright 2017 Ning Xu"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/99515","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Huang, Thomas","Hasegawa-Johnson, Mark","Lazebnik, Svetlana","Liang, Zhi-Pei"]},{"key":"dc:creator","label":"Author","values":["Xu, Ning"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2018-03-13T17:35:49Z","2020-03-14T09:15:22Z","2017-12-06","2017-12"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical & Computer Engr"]},{"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":["Video segmentation","Object selection","Computer vision","Deep learning","Image segmentation"]}]},{"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 Ning Xu"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/99515"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["\"Image and video object selection present fundamental research problems in the computer vision field and have many practical applications. They are important technologies in image and video editing, film production, robotics and autonomous driving etc. Previous methods have serious limitations for those tasks for several reasons. First, most of them use some low-level, handcrafted features which are not optimal. Second, they also lack the high-level understanding of \"\"objectness\"\" and semantics. Last but not the least, their generalization ability on unconstrained scenarios is very poor. Recently, deep learning has become the dominant method for computer vision tasks including recognition and detection since it cannot only learn good feature representation in an end-to-end manner but it is also effective at capturing the high-level semantics. However, its exploration in image and video object selection is still impoverished. Therefore, in this thesis we propose several novel deep-learning based methods to tackle the limitations in image and video object selection. Our algorithms are easy to understand and effective. Experimental results clearly demonstrate the superiority of our algorithms over previous methods. Some highlights include the following: (1) Our interactive segmentation algorithm is the first deep-learning based algorithm and achieves the state-of-the-art results on both small-scale and large-scale benchmarks. (2) Our rectangle-based algorithm novelly transforms rectangle inputs to attention-like distance maps and achieves robust performance for sloppy user selections or misplaced detection boxes. (3) Our image matting algorithm is the first to demonstrate the feasibility of learning an alpha matte end-to-end given an image and trimap. It also achieves state-of-the-art results on image matting and video matting benchmarks. (4) Our video object segmentation method combines CNN network with RNN memory cells to learn both good image feature representation and the temporal-spatial coherence.\"","Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2019-12-01","The student, Ning Xu, accepted the attached license on 2017-12-05 at 15:16.","The student, Ning Xu, submitted this Dissertation for approval on 2017-12-05 at 15:31.","This Dissertation was approved for publication on 2017-12-06 at 15:36.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11868 on 2018-03-13 at 10:37:42","Made available in DSpace on 2018-03-13T17:35:49Z (GMT). No. of bitstreams: 3 XU-DISSERTATION-2017.pdf: 11792512 bytes, checksum: 6b67db02c6f837d007a99ff3afb78b69 (MD5) LICENSE.txt: 4204 bytes, checksum: a5d170e0dde552cea19909deae515a09 (MD5) PROQUEST_LICENSE.txt: 4550 bytes, checksum: d9b4c449de418eb48b4d25ceb1cd2da3 (MD5) Previous issue date: 2017-12-06","Embargo set by: Seth Robbins for item 105484 Lift date: 2020-03-13T17:36:05Z Reason: Author requested closed access (OA after 2yrs) in Vireo ETD system","Limited Restriction Lifted for Item 105484 on 2020-03-14T09:15:22Z."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Image and video object selection"]}]}],"canonical_facts":{"dc:contributor":["Huang, Thomas","Hasegawa-Johnson, Mark","Lazebnik, Svetlana","Liang, Zhi-Pei"],"dc:creator":["Xu, Ning"],"dc:date":["2018-03-13T17:35:49Z","2020-03-14T09:15:22Z","2017-12-06","2017-12"],"dc:description":["\"Image and video object selection present fundamental research problems in the computer vision field and have many practical applications. They are important technologies in image and video editing, film production, robotics and autonomous driving etc. Previous methods have serious limitations for those tasks for several reasons. First, most of them use some low-level, handcrafted features which are not optimal. Second, they also lack the high-level understanding of \"\"objectness\"\" and semantics. Last but not the least, their generalization ability on unconstrained scenarios is very poor. Recently, deep learning has become the dominant method for computer vision tasks including recognition and detection since it cannot only learn good feature representation in an end-to-end manner but it is also effective at capturing the high-level semantics. However, its exploration in image and video object selection is still impoverished. Therefore, in this thesis we propose several novel deep-learning based methods to tackle the limitations in image and video object selection. Our algorithms are easy to understand and effective. Experimental results clearly demonstrate the superiority of our algorithms over previous methods. Some highlights include the following: (1) Our interactive segmentation algorithm is the first deep-learning based algorithm and achieves the state-of-the-art results on both small-scale and large-scale benchmarks. (2) Our rectangle-based algorithm novelly transforms rectangle inputs to attention-like distance maps and achieves robust performance for sloppy user selections or misplaced detection boxes. (3) Our image matting algorithm is the first to demonstrate the feasibility of learning an alpha matte end-to-end given an image and trimap. It also achieves state-of-the-art results on image matting and video matting benchmarks. (4) Our video object segmentation method combines CNN network with RNN memory cells to learn both good image feature representation and the temporal-spatial coherence.\"","Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2019-12-01","The student, Ning Xu, accepted the attached license on 2017-12-05 at 15:16.","The student, Ning Xu, submitted this Dissertation for approval on 2017-12-05 at 15:31.","This Dissertation was approved for publication on 2017-12-06 at 15:36.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11868 on 2018-03-13 at 10:37:42","Made available in DSpace on 2018-03-13T17:35:49Z (GMT). No. of bitstreams: 3 XU-DISSERTATION-2017.pdf: 11792512 bytes, checksum: 6b67db02c6f837d007a99ff3afb78b69 (MD5) LICENSE.txt: 4204 bytes, checksum: a5d170e0dde552cea19909deae515a09 (MD5) PROQUEST_LICENSE.txt: 4550 bytes, checksum: d9b4c449de418eb48b4d25ceb1cd2da3 (MD5) Previous issue date: 2017-12-06","Embargo set by: Seth Robbins for item 105484 Lift date: 2020-03-13T17:36:05Z Reason: Author requested closed access (OA after 2yrs) in Vireo ETD system","Limited Restriction Lifted for Item 105484 on 2020-03-14T09:15:22Z."],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/99515"],"dc:language":["en"],"dc:rights":["Copyright 2017 Ning Xu"],"dc:subject":["Video segmentation","Object selection","Computer vision","Deep learning","Image segmentation"],"dc:title":["Image and video object selection"],"dc:type":["text"],"thesis:degree_discipline":["Electrical & Computer Engr"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:24:37Z"}