Predicting emotional states of images using Bayesian multiple kernel learning

He Zhang, Mehmet Gonen, Zhirong Yang, Erkki Oja

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

2 Citations (Scopus)

Abstract

Images usually convey information that can influence people's emotional states. Such affective information can be used by search engines and social networks for better understanding the user's preferences. We propose here a novel Bayesian multiple kernel learning method for predicting the emotions evoked by images. The proposed method can make use of different image features simultaneously to obtain a better prediction performance, with the advantage of automatically selecting important features. Specifically, our method has been implemented within a multilabel setup in order to capture the correlations between emotions. Due to its probabilistic nature, our method is also able to produce probabilistic outputs for measuring a distribution of emotional intensities. The experimental results on the International Affective Picture System (IAPS) dataset show that the proposed approach achieves a bette classification performance and provides a more interpretable feature selection capability than the state-of-the-art methods.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages274-282
Number of pages9
Volume8228 LNCS
EditionPART 3
DOIs
StatePublished - 2013
Externally publishedYes
Event20th International Conference on Neural Information Processing, ICONIP 2013 - Daegu, Korea, Republic of
Duration: Nov 3 2013Nov 7 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume8228 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other20th International Conference on Neural Information Processing, ICONIP 2013
CountryKorea, Republic of
CityDaegu
Period11/3/1311/7/13

Fingerprint

Search engines
Feature extraction
kernel
User Preferences
Performance Prediction
Search Engine
Feature Selection
Social Networks
Emotion
Learning
Output
Experimental Results

Keywords

  • Image emotion
  • Low-level image features
  • Multiple kernel learning
  • Multiview learning
  • Variational approximation

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Zhang, H., Gonen, M., Yang, Z., & Oja, E. (2013). Predicting emotional states of images using Bayesian multiple kernel learning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 3 ed., Vol. 8228 LNCS, pp. 274-282). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8228 LNCS, No. PART 3). https://doi.org/10.1007/978-3-642-42051-1_35

Predicting emotional states of images using Bayesian multiple kernel learning. / Zhang, He; Gonen, Mehmet; Yang, Zhirong; Oja, Erkki.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8228 LNCS PART 3. ed. 2013. p. 274-282 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8228 LNCS, No. PART 3).

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

Zhang, H, Gonen, M, Yang, Z & Oja, E 2013, Predicting emotional states of images using Bayesian multiple kernel learning. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 edn, vol. 8228 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 8228 LNCS, pp. 274-282, 20th International Conference on Neural Information Processing, ICONIP 2013, Daegu, Korea, Republic of, 11/3/13. https://doi.org/10.1007/978-3-642-42051-1_35
Zhang H, Gonen M, Yang Z, Oja E. Predicting emotional states of images using Bayesian multiple kernel learning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 ed. Vol. 8228 LNCS. 2013. p. 274-282. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). https://doi.org/10.1007/978-3-642-42051-1_35
Zhang, He ; Gonen, Mehmet ; Yang, Zhirong ; Oja, Erkki. / Predicting emotional states of images using Bayesian multiple kernel learning. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8228 LNCS PART 3. ed. 2013. pp. 274-282 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{c4c98d4c823b41148e44f1124acab598,
title = "Predicting emotional states of images using Bayesian multiple kernel learning",
abstract = "Images usually convey information that can influence people's emotional states. Such affective information can be used by search engines and social networks for better understanding the user's preferences. We propose here a novel Bayesian multiple kernel learning method for predicting the emotions evoked by images. The proposed method can make use of different image features simultaneously to obtain a better prediction performance, with the advantage of automatically selecting important features. Specifically, our method has been implemented within a multilabel setup in order to capture the correlations between emotions. Due to its probabilistic nature, our method is also able to produce probabilistic outputs for measuring a distribution of emotional intensities. The experimental results on the International Affective Picture System (IAPS) dataset show that the proposed approach achieves a bette classification performance and provides a more interpretable feature selection capability than the state-of-the-art methods.",
keywords = "Image emotion, Low-level image features, Multiple kernel learning, Multiview learning, Variational approximation",
author = "He Zhang and Mehmet Gonen and Zhirong Yang and Erkki Oja",
year = "2013",
doi = "10.1007/978-3-642-42051-1_35",
language = "English (US)",
isbn = "9783642420504",
volume = "8228 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 3",
pages = "274--282",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 3",

}

TY - GEN

T1 - Predicting emotional states of images using Bayesian multiple kernel learning

AU - Zhang, He

AU - Gonen, Mehmet

AU - Yang, Zhirong

AU - Oja, Erkki

PY - 2013

Y1 - 2013

N2 - Images usually convey information that can influence people's emotional states. Such affective information can be used by search engines and social networks for better understanding the user's preferences. We propose here a novel Bayesian multiple kernel learning method for predicting the emotions evoked by images. The proposed method can make use of different image features simultaneously to obtain a better prediction performance, with the advantage of automatically selecting important features. Specifically, our method has been implemented within a multilabel setup in order to capture the correlations between emotions. Due to its probabilistic nature, our method is also able to produce probabilistic outputs for measuring a distribution of emotional intensities. The experimental results on the International Affective Picture System (IAPS) dataset show that the proposed approach achieves a bette classification performance and provides a more interpretable feature selection capability than the state-of-the-art methods.

AB - Images usually convey information that can influence people's emotional states. Such affective information can be used by search engines and social networks for better understanding the user's preferences. We propose here a novel Bayesian multiple kernel learning method for predicting the emotions evoked by images. The proposed method can make use of different image features simultaneously to obtain a better prediction performance, with the advantage of automatically selecting important features. Specifically, our method has been implemented within a multilabel setup in order to capture the correlations between emotions. Due to its probabilistic nature, our method is also able to produce probabilistic outputs for measuring a distribution of emotional intensities. The experimental results on the International Affective Picture System (IAPS) dataset show that the proposed approach achieves a bette classification performance and provides a more interpretable feature selection capability than the state-of-the-art methods.

KW - Image emotion

KW - Low-level image features

KW - Multiple kernel learning

KW - Multiview learning

KW - Variational approximation

UR - http://www.scopus.com/inward/record.url?scp=84893414123&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893414123&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-42051-1_35

DO - 10.1007/978-3-642-42051-1_35

M3 - Conference contribution

SN - 9783642420504

VL - 8228 LNCS

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

SP - 274

EP - 282

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

ER -