Using multiple segmentations to discover objects and their extent in image collections
Given a large dataset of images, we seek to automatically determine the visually similar object and scene classes together with their image segmentation. To achieve this we combine two ideas: (i) that a set of segmented objects can be partitioned into visual object classes using topic discovery mode...
| Main Authors: | , , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
IEEE
2006
|
| _version_ | 1824458974260363264 |
|---|---|
| author | Russell, BC Efros, AA Sivic, J Freeman, WT Zisserman, A |
| author_facet | Russell, BC Efros, AA Sivic, J Freeman, WT Zisserman, A |
| author_sort | Russell, BC |
| collection | OXFORD |
| description | Given a large dataset of images, we seek to automatically determine the visually similar object and scene classes together with their image segmentation. To achieve this we combine two ideas: (i) that a set of segmented objects can be partitioned into visual object classes using topic discovery models from statistical text analysis; and (ii) that visual object classes can be used to assess the accuracy of a segmentation. To tie these ideas together we compute multiple segmentations of each image and then: (i) learn the object classes; and (ii) choose the correct segmentations. We demonstrate that such an algorithm succeeds in automatically discovering many familiar objects in a variety of image datasets, including those from Caltech, MSRC and LabelMe. |
| first_indexed | 2025-02-19T04:34:25Z |
| format | Conference item |
| id | oxford-uuid:332ff331-d7e6-48a9-88c1-043975488194 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2025-02-19T04:34:25Z |
| publishDate | 2006 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:332ff331-d7e6-48a9-88c1-0439754881942025-01-24T15:50:04ZUsing multiple segmentations to discover objects and their extent in image collectionsConference itemhttp://purl.org/coar/resource_type/c_5794uuid:332ff331-d7e6-48a9-88c1-043975488194EnglishSymplectic ElementsIEEE2006Russell, BCEfros, AASivic, JFreeman, WTZisserman, AGiven a large dataset of images, we seek to automatically determine the visually similar object and scene classes together with their image segmentation. To achieve this we combine two ideas: (i) that a set of segmented objects can be partitioned into visual object classes using topic discovery models from statistical text analysis; and (ii) that visual object classes can be used to assess the accuracy of a segmentation. To tie these ideas together we compute multiple segmentations of each image and then: (i) learn the object classes; and (ii) choose the correct segmentations. We demonstrate that such an algorithm succeeds in automatically discovering many familiar objects in a variety of image datasets, including those from Caltech, MSRC and LabelMe. |
| spellingShingle | Russell, BC Efros, AA Sivic, J Freeman, WT Zisserman, A Using multiple segmentations to discover objects and their extent in image collections |
| title | Using multiple segmentations to discover objects and their extent in image collections |
| title_full | Using multiple segmentations to discover objects and their extent in image collections |
| title_fullStr | Using multiple segmentations to discover objects and their extent in image collections |
| title_full_unstemmed | Using multiple segmentations to discover objects and their extent in image collections |
| title_short | Using multiple segmentations to discover objects and their extent in image collections |
| title_sort | using multiple segmentations to discover objects and their extent in image collections |
| work_keys_str_mv | AT russellbc usingmultiplesegmentationstodiscoverobjectsandtheirextentinimagecollections AT efrosaa usingmultiplesegmentationstodiscoverobjectsandtheirextentinimagecollections AT sivicj usingmultiplesegmentationstodiscoverobjectsandtheirextentinimagecollections AT freemanwt usingmultiplesegmentationstodiscoverobjectsandtheirextentinimagecollections AT zissermana usingmultiplesegmentationstodiscoverobjectsandtheirextentinimagecollections |