Identifying emotions aroused from paintings

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Understanding the emotional appeal of paintings is a significant research problem related to affective image classification. The problem is challenging in part due to the scarceness of manually-classified paintings. Our work proposes to apply statistical models trained over photographs to infer the emotional appeal of paintings. Directly applying the learned models on photographs to paintings cannot provide accurate classification results, because visual features extracted from paintings and natural photographs have different characteristics. This work presents an adaptive learning algorithm that leverages labeled photographs and unlabeled paintings to infer the visual appeal of paintings. In particular, we iteratively adapt the feature distribution in photographs to fit paintings and maximize the joint likelihood of labeled and unlabeled data. We evaluate our approach through two emotional classification tasks: distinguishing positive from negative emotions, and differentiating reactive emotions from non-reactive ones. Experimental results show the potential of our approach.

Original languageEnglish (US)
Title of host publicationComputer Vision - ECCV 2016 Workshops, Proceedings
EditorsGang Hua, Hervé Jégou
PublisherSpringer Verlag
Pages48-63
Number of pages16
ISBN (Print)9783319466033
DOIs
StatePublished - Jan 1 2016
Event14th European Conference on Computer Vision, ECCV 2016 - Amsterdam, Netherlands
Duration: Oct 8 2016Oct 16 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9913 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other14th European Conference on Computer Vision, ECCV 2016
CountryNetherlands
CityAmsterdam
Period10/8/1610/16/16

Fingerprint

Painting
Appeal
Adaptive Learning
Emotion
Image classification
Image Classification
Adaptive algorithms
Adaptive Algorithm
Leverage
Learning algorithms
Statistical Model
Learning Algorithm
Likelihood
Maximise
Evaluate
Experimental Results

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Lu, X., Sawant, N., Newman, M. G., Adams, R. B., Wang, J. Z., & Li, J. (2016). Identifying emotions aroused from paintings. In G. Hua, & H. Jégou (Eds.), Computer Vision - ECCV 2016 Workshops, Proceedings (pp. 48-63). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9913 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46604-0_4
Lu, Xin ; Sawant, Neela ; Newman, Michelle G. ; Adams, Reginald B. ; Wang, James Z. ; Li, Jia. / Identifying emotions aroused from paintings. Computer Vision - ECCV 2016 Workshops, Proceedings. editor / Gang Hua ; Hervé Jégou. Springer Verlag, 2016. pp. 48-63 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{9eff2e8c04404a398f82f209f6c7c138,
title = "Identifying emotions aroused from paintings",
abstract = "Understanding the emotional appeal of paintings is a significant research problem related to affective image classification. The problem is challenging in part due to the scarceness of manually-classified paintings. Our work proposes to apply statistical models trained over photographs to infer the emotional appeal of paintings. Directly applying the learned models on photographs to paintings cannot provide accurate classification results, because visual features extracted from paintings and natural photographs have different characteristics. This work presents an adaptive learning algorithm that leverages labeled photographs and unlabeled paintings to infer the visual appeal of paintings. In particular, we iteratively adapt the feature distribution in photographs to fit paintings and maximize the joint likelihood of labeled and unlabeled data. We evaluate our approach through two emotional classification tasks: distinguishing positive from negative emotions, and differentiating reactive emotions from non-reactive ones. Experimental results show the potential of our approach.",
author = "Xin Lu and Neela Sawant and Newman, {Michelle G.} and Adams, {Reginald B.} and Wang, {James Z.} and Jia Li",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-3-319-46604-0_4",
language = "English (US)",
isbn = "9783319466033",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "48--63",
editor = "Gang Hua and Herv{\'e} J{\'e}gou",
booktitle = "Computer Vision - ECCV 2016 Workshops, Proceedings",
address = "Germany",

}

Lu, X, Sawant, N, Newman, MG, Adams, RB, Wang, JZ & Li, J 2016, Identifying emotions aroused from paintings. in G Hua & H Jégou (eds), Computer Vision - ECCV 2016 Workshops, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9913 LNCS, Springer Verlag, pp. 48-63, 14th European Conference on Computer Vision, ECCV 2016, Amsterdam, Netherlands, 10/8/16. https://doi.org/10.1007/978-3-319-46604-0_4

Identifying emotions aroused from paintings. / Lu, Xin; Sawant, Neela; Newman, Michelle G.; Adams, Reginald B.; Wang, James Z.; Li, Jia.

Computer Vision - ECCV 2016 Workshops, Proceedings. ed. / Gang Hua; Hervé Jégou. Springer Verlag, 2016. p. 48-63 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9913 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Identifying emotions aroused from paintings

AU - Lu, Xin

AU - Sawant, Neela

AU - Newman, Michelle G.

AU - Adams, Reginald B.

AU - Wang, James Z.

AU - Li, Jia

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Understanding the emotional appeal of paintings is a significant research problem related to affective image classification. The problem is challenging in part due to the scarceness of manually-classified paintings. Our work proposes to apply statistical models trained over photographs to infer the emotional appeal of paintings. Directly applying the learned models on photographs to paintings cannot provide accurate classification results, because visual features extracted from paintings and natural photographs have different characteristics. This work presents an adaptive learning algorithm that leverages labeled photographs and unlabeled paintings to infer the visual appeal of paintings. In particular, we iteratively adapt the feature distribution in photographs to fit paintings and maximize the joint likelihood of labeled and unlabeled data. We evaluate our approach through two emotional classification tasks: distinguishing positive from negative emotions, and differentiating reactive emotions from non-reactive ones. Experimental results show the potential of our approach.

AB - Understanding the emotional appeal of paintings is a significant research problem related to affective image classification. The problem is challenging in part due to the scarceness of manually-classified paintings. Our work proposes to apply statistical models trained over photographs to infer the emotional appeal of paintings. Directly applying the learned models on photographs to paintings cannot provide accurate classification results, because visual features extracted from paintings and natural photographs have different characteristics. This work presents an adaptive learning algorithm that leverages labeled photographs and unlabeled paintings to infer the visual appeal of paintings. In particular, we iteratively adapt the feature distribution in photographs to fit paintings and maximize the joint likelihood of labeled and unlabeled data. We evaluate our approach through two emotional classification tasks: distinguishing positive from negative emotions, and differentiating reactive emotions from non-reactive ones. Experimental results show the potential of our approach.

UR - http://www.scopus.com/inward/record.url?scp=84989820787&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84989820787&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-46604-0_4

DO - 10.1007/978-3-319-46604-0_4

M3 - Conference contribution

AN - SCOPUS:84989820787

SN - 9783319466033

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 48

EP - 63

BT - Computer Vision - ECCV 2016 Workshops, Proceedings

A2 - Hua, Gang

A2 - Jégou, Hervé

PB - Springer Verlag

ER -

Lu X, Sawant N, Newman MG, Adams RB, Wang JZ, Li J. Identifying emotions aroused from paintings. In Hua G, Jégou H, editors, Computer Vision - ECCV 2016 Workshops, Proceedings. Springer Verlag. 2016. p. 48-63. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-46604-0_4