Efficient retrieval of deformable shape classes using local self-similarities

We present an efficient object retrieval system based on the identification of abstract deformable 'shape' classes using the self-similarity descriptor of Shechtman and Irani. Given a user-specified query object, we retrieve other images which share a common `shape' even if their appe...

Full description

Bibliographic Details
Main Authors: Chatfield, K, Philbin, J, Zisserman, A
Format: Conference item
Language:English
Published: IEEE 2010
_version_ 1824458871140253696
author Chatfield, K
Philbin, J
Zisserman, A
author_facet Chatfield, K
Philbin, J
Zisserman, A
author_sort Chatfield, K
collection OXFORD
description We present an efficient object retrieval system based on the identification of abstract deformable 'shape' classes using the self-similarity descriptor of Shechtman and Irani. Given a user-specified query object, we retrieve other images which share a common `shape' even if their appearance differs greatly in terms of colour, texture, edges and other common photometric properties. In order to use the self-similarity descriptor for efficient retrieval we make three contributions: (i) we sparsify the descriptor points by locating discriminative regions within each image, thus reducing the computational expense of shape matching; (ii) we extend to enable matching despite changes in scale; and (iii) we show that vector quantizing the descriptor does not inhibit performance, thus providing the basis of a large-scale shape-based retrieval system using a bag-of-visual-words approach. Performance is demonstrated on the challenging ETHZ deformable shape dataset and a full episode from the television series Lost, and is shown to be superior to appearance-based approaches for matching non-rigid shape classes.
first_indexed 2025-02-19T04:32:46Z
format Conference item
id oxford-uuid:2d59918a-4579-4aec-83fa-bb672f1768c8
institution University of Oxford
language English
last_indexed 2025-02-19T04:32:46Z
publishDate 2010
publisher IEEE
record_format dspace
spelling oxford-uuid:2d59918a-4579-4aec-83fa-bb672f1768c82025-01-15T13:26:35ZEfficient retrieval of deformable shape classes using local self-similaritiesConference itemhttp://purl.org/coar/resource_type/c_5794uuid:2d59918a-4579-4aec-83fa-bb672f1768c8EnglishSymplectic ElementsIEEE2010Chatfield, KPhilbin, JZisserman, AWe present an efficient object retrieval system based on the identification of abstract deformable 'shape' classes using the self-similarity descriptor of Shechtman and Irani. Given a user-specified query object, we retrieve other images which share a common `shape' even if their appearance differs greatly in terms of colour, texture, edges and other common photometric properties. In order to use the self-similarity descriptor for efficient retrieval we make three contributions: (i) we sparsify the descriptor points by locating discriminative regions within each image, thus reducing the computational expense of shape matching; (ii) we extend to enable matching despite changes in scale; and (iii) we show that vector quantizing the descriptor does not inhibit performance, thus providing the basis of a large-scale shape-based retrieval system using a bag-of-visual-words approach. Performance is demonstrated on the challenging ETHZ deformable shape dataset and a full episode from the television series Lost, and is shown to be superior to appearance-based approaches for matching non-rigid shape classes.
spellingShingle Chatfield, K
Philbin, J
Zisserman, A
Efficient retrieval of deformable shape classes using local self-similarities
title Efficient retrieval of deformable shape classes using local self-similarities
title_full Efficient retrieval of deformable shape classes using local self-similarities
title_fullStr Efficient retrieval of deformable shape classes using local self-similarities
title_full_unstemmed Efficient retrieval of deformable shape classes using local self-similarities
title_short Efficient retrieval of deformable shape classes using local self-similarities
title_sort efficient retrieval of deformable shape classes using local self similarities
work_keys_str_mv AT chatfieldk efficientretrievalofdeformableshapeclassesusinglocalselfsimilarities
AT philbinj efficientretrievalofdeformableshapeclassesusinglocalselfsimilarities
AT zissermana efficientretrievalofdeformableshapeclassesusinglocalselfsimilarities