{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/98355"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/98355","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Object localization in natural images","abstract":"\"Object localization algorithms aim at finding out what objects exist in an image and where each object is. Object localization is fundamental to many computer vision problems. Simply knowing what is in the image is not enough when we want to reason about object properties such as shape, color or pose, and infer the relationship between different objects. Object localization is also a challenging task. Intra-class variance as well as differences in illumination, viewing distance and viewpoint all make objects within the same category look different from each other. The goal of this thesis is to study the problem of object localization in natural images. We start off by answering the three key questions about object localization: \"\"how\"\" good are our algorithms at localizing objects, \"\"what\"\" are the types of mistakes our algorithms commonly make, and \"\"why\"\" are certain objects hard to detect by providing a set of annotations and tools to facilitate analysis on object localization algorithms. One of the conclusions drawn from our analysis is that object detectors sometimes succeed at finding the rough location of the object but fail to precisely locate the object. Inspired by this observation, we proposed a framework that takes as input a roughly localized object bounding box, and generates a more accurate localization of the object, both in the form of a bounding box and pixel-wise segmentation. Finally, we try to explore an alternative way to localize objects, by detecting object boundaries, which could generalize to localizing objects of unknown categories.\"","abstract_html":""Object localization algorithms aim at finding out what objects exist in an image and where each object is. Object localization is fundamental to many computer vision problems. Simply knowing what is in the image is not enough when we want to reason about object properties such as shape, color or pose, and infer the relationship between different objects. Object localization is also a challenging task. Intra-class variance as well as differences in illumination, viewing distance and viewpoint all make objects within the same category look different from each other. The goal of this thesis is to study the problem of object localization in natural images. We start off by answering the three key questions about object localization: ""how"" good are our algorithms at localizing objects, ""what"" are the types of mistakes our algorithms commonly make, and ""why"" are certain objects hard to detect by providing a set of annotations and tools to facilitate analysis on object localization algorithms. One of the conclusions drawn from our analysis is that object detectors sometimes succeed at finding the rough location of the object but fail to precisely locate the object. Inspired by this observation, we proposed a framework that takes as input a roughly localized object bounding box, and generates a more accurate localization of the object, both in the form of a bounding box and pixel-wise segmentation. Finally, we try to explore an alternative way to localize objects, by detecting object boundaries, which could generalize to localizing objects of unknown categories."","abstract_has_math":false,"creators":["Dai, Qieyun"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Hoiem, Derek","Forsyth, David","Lazebnik, Svetlana","Ramanan, Deva"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2017,"date_issued":"2017-09-29T17:56:38Z","date_published":"2017-09-29T17:56:38Z","updated_at":"2026-07-22T22:24:35Z","subjects":["Object localization","Detector performance analysis","Semantic segmentation","Structured learning"],"languages":["en"],"rights":["Copyright 2017 Qieyun Dai"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/98355","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Hoiem, Derek","Forsyth, David","Lazebnik, Svetlana","Ramanan, Deva"]},{"key":"dc:creator","label":"Author","values":["Dai, Qieyun"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2017-09-29T17:56:38Z","2017-07-11","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":["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":["Object localization","Detector performance analysis","Semantic segmentation","Structured learning"]}]},{"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 Qieyun Dai"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/98355"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["\"Object localization algorithms aim at finding out what objects exist in an image and where each object is. Object localization is fundamental to many computer vision problems. Simply knowing what is in the image is not enough when we want to reason about object properties such as shape, color or pose, and infer the relationship between different objects. Object localization is also a challenging task. Intra-class variance as well as differences in illumination, viewing distance and viewpoint all make objects within the same category look different from each other. The goal of this thesis is to study the problem of object localization in natural images. We start off by answering the three key questions about object localization: \"\"how\"\" good are our algorithms at localizing objects, \"\"what\"\" are the types of mistakes our algorithms commonly make, and \"\"why\"\" are certain objects hard to detect by providing a set of annotations and tools to facilitate analysis on object localization algorithms. One of the conclusions drawn from our analysis is that object detectors sometimes succeed at finding the rough location of the object but fail to precisely locate the object. Inspired by this observation, we proposed a framework that takes as input a roughly localized object bounding box, and generates a more accurate localization of the object, both in the form of a bounding box and pixel-wise segmentation. Finally, we try to explore an alternative way to localize objects, by detecting object boundaries, which could generalize to localizing objects of unknown categories.\"","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2017-09-29 without embargo terms","The student, Qieyun Dai, accepted the attached license on 2017-07-10 at 10:29.","The student, Qieyun Dai, submitted this Dissertation for approval on 2017-07-10 at 10:47.","This Dissertation was approved for publication on 2017-07-11 at 08:20.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11359 on 2017-09-29 at 11:29:10","Made available in DSpace on 2017-09-29T17:56:38Z (GMT). No. of bitstreams: 3 DAI-DISSERTATION-2017.pdf: 93230782 bytes, checksum: ec3a89ac81d21cbf7d4dd94f730e2bf3 (MD5) LICENSE.txt: 4207 bytes, checksum: 50b11875a580dd545c67055669a75c03 (MD5) PROQUEST_LICENSE.txt: 4553 bytes, checksum: e5c6203e144121ac4c111c92cac1d4f5 (MD5) Previous issue date: 2017-07-11"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Object localization in natural images"]}]}],"canonical_facts":{"dc:contributor":["Hoiem, Derek","Forsyth, David","Lazebnik, Svetlana","Ramanan, Deva"],"dc:creator":["Dai, Qieyun"],"dc:date":["2017-09-29T17:56:38Z","2017-07-11","2017-08"],"dc:description":["\"Object localization algorithms aim at finding out what objects exist in an image and where each object is. Object localization is fundamental to many computer vision problems. Simply knowing what is in the image is not enough when we want to reason about object properties such as shape, color or pose, and infer the relationship between different objects. Object localization is also a challenging task. Intra-class variance as well as differences in illumination, viewing distance and viewpoint all make objects within the same category look different from each other. The goal of this thesis is to study the problem of object localization in natural images. We start off by answering the three key questions about object localization: \"\"how\"\" good are our algorithms at localizing objects, \"\"what\"\" are the types of mistakes our algorithms commonly make, and \"\"why\"\" are certain objects hard to detect by providing a set of annotations and tools to facilitate analysis on object localization algorithms. One of the conclusions drawn from our analysis is that object detectors sometimes succeed at finding the rough location of the object but fail to precisely locate the object. Inspired by this observation, we proposed a framework that takes as input a roughly localized object bounding box, and generates a more accurate localization of the object, both in the form of a bounding box and pixel-wise segmentation. Finally, we try to explore an alternative way to localize objects, by detecting object boundaries, which could generalize to localizing objects of unknown categories.\"","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2017-09-29 without embargo terms","The student, Qieyun Dai, accepted the attached license on 2017-07-10 at 10:29.","The student, Qieyun Dai, submitted this Dissertation for approval on 2017-07-10 at 10:47.","This Dissertation was approved for publication on 2017-07-11 at 08:20.","DSpace SAF Submission Ingestion Package generated from Vireo submission #11359 on 2017-09-29 at 11:29:10","Made available in DSpace on 2017-09-29T17:56:38Z (GMT). No. of bitstreams: 3 DAI-DISSERTATION-2017.pdf: 93230782 bytes, checksum: ec3a89ac81d21cbf7d4dd94f730e2bf3 (MD5) LICENSE.txt: 4207 bytes, checksum: 50b11875a580dd545c67055669a75c03 (MD5) PROQUEST_LICENSE.txt: 4553 bytes, checksum: e5c6203e144121ac4c111c92cac1d4f5 (MD5) Previous issue date: 2017-07-11"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/98355"],"dc:language":["en"],"dc:rights":["Copyright 2017 Qieyun Dai"],"dc:subject":["Object localization","Detector performance analysis","Semantic segmentation","Structured learning"],"dc:title":["Object localization in natural images"],"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":"2026-07-22T22:24:35Z"}