Face painting: querying art with photos

We study the problem of matching photos of a person to paintings of that person, in order to retrieve similar paintings given a query photo. This is challenging as paintings span many media (oil, ink, watercolor) and can vary tremendously in style (caricature, pop art, minimalist). We make the follo...

Full description

Bibliographic Details
Main Authors: Crowley, EJ, Parkhi, OM, Zisserman, A
Format: Conference item
Language:English
Published: British Machine Vision Association 2015
_version_ 1826315168223592448
author Crowley, EJ
Parkhi, OM
Zisserman, A
author_facet Crowley, EJ
Parkhi, OM
Zisserman, A
author_sort Crowley, EJ
collection OXFORD
description We study the problem of matching photos of a person to paintings of that person, in order to retrieve similar paintings given a query photo. This is challenging as paintings span many media (oil, ink, watercolor) and can vary tremendously in style (caricature, pop art, minimalist). We make the following contributions: (i) we show that, depending on the face representation used, performance can be improved substantially by learning -- either by a linear projection matrix common across identities, or by a per-identity classifier. We compare Fisher Vector and Convolutional Neural Network representations for this task; (ii) we introduce new datasets for learning and evaluating this problem; (iii) we also consider the reverse problem of retrieving photos from a large corpus given a painting; and finally, (iv) using the learnt descriptors, we show that, given a photo of a person, we are able to find their doppelgänger in a large dataset of oil paintings, and how this result can be varied by modifying attributes (e.g. frowning, old looking).
first_indexed 2024-12-09T03:20:45Z
format Conference item
id oxford-uuid:c9f058a7-a2d8-4edd-b351-91ea81f4e582
institution University of Oxford
language English
last_indexed 2024-12-09T03:20:45Z
publishDate 2015
publisher British Machine Vision Association
record_format dspace
spelling oxford-uuid:c9f058a7-a2d8-4edd-b351-91ea81f4e5822024-11-05T14:32:42ZFace painting: querying art with photosConference itemhttp://purl.org/coar/resource_type/c_5794uuid:c9f058a7-a2d8-4edd-b351-91ea81f4e582EnglishSymplectic ElementsBritish Machine Vision Association2015Crowley, EJParkhi, OMZisserman, AWe study the problem of matching photos of a person to paintings of that person, in order to retrieve similar paintings given a query photo. This is challenging as paintings span many media (oil, ink, watercolor) and can vary tremendously in style (caricature, pop art, minimalist). We make the following contributions: (i) we show that, depending on the face representation used, performance can be improved substantially by learning -- either by a linear projection matrix common across identities, or by a per-identity classifier. We compare Fisher Vector and Convolutional Neural Network representations for this task; (ii) we introduce new datasets for learning and evaluating this problem; (iii) we also consider the reverse problem of retrieving photos from a large corpus given a painting; and finally, (iv) using the learnt descriptors, we show that, given a photo of a person, we are able to find their doppelgänger in a large dataset of oil paintings, and how this result can be varied by modifying attributes (e.g. frowning, old looking).
spellingShingle Crowley, EJ
Parkhi, OM
Zisserman, A
Face painting: querying art with photos
title Face painting: querying art with photos
title_full Face painting: querying art with photos
title_fullStr Face painting: querying art with photos
title_full_unstemmed Face painting: querying art with photos
title_short Face painting: querying art with photos
title_sort face painting querying art with photos
work_keys_str_mv AT crowleyej facepaintingqueryingartwithphotos
AT parkhiom facepaintingqueryingartwithphotos
AT zissermana facepaintingqueryingartwithphotos