Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion

Khiet Phuong Truong; David A. van Leeuwen; Mark A. Neerincx; Franciska M.G. de Jong

Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion

Khiet Phuong Truong, David A. van Leeuwen, Mark A. Neerincx, Franciska M.G. de Jong

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

21 Citations (Scopus)

204 Downloads (Pure)

Abstract

In this paper, we describe emotion recognition experiments carried out for spontaneous aﬀective speech with the aim to compare the added value of annotation of felt emotion versus annotation of perceived emotion. Using speech material available in the TNO-GAMING corpus (a corpus containing audiovisual recordings of people playing videogames), speech-based aﬀect recognizers were developed that can predict Arousal and Valence scalar values. Two types of recognizers were developed in parallel: one trained with felt emotion annotations (generated by the gamers themselves) and one trained with perceived/observed emotion annotations (generated by a group of observers). The experiments showed that, in speech, with the methods and features currently used, observed emotions are easier to predict than felt emotions. The results suggest that recognition performance strongly depends on how and by whom the emotion annotations are carried out.

Original language	English
Title of host publication	Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009)
Publisher	International Speech Communication Association
Pages	2027-2030
Number of pages	4
Publication status	Published - 2009
Event	10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009 - Brighton, United Kingdom Duration: 6 Sept 2009 → 10 Sept 2009 Conference number: 10 http://www.interspeech2010.jpn.org/

Publication series

Name	Publications Speech Processing Group, Brno University of Technology
Publisher	International Speech Communication Association
ISSN (Print)	1990-9772

Conference

Conference	10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009
Abbreviated title	INTERSPEECH
Country/Territory	United Kingdom
City	Brighton
Period	6/09/09 → 10/09/09
Internet address	http://www.interspeech2010.jpn.org/

Keywords

IR-68948
EWI-17024
METIS-264250

Access to Document

Truongs2009arousal_submitted_v2Submitted version, 68 KB

Cite this

Truong, K. P., van Leeuwen, D. A., Neerincx, M. A., & de Jong, F. M. G. (2009). Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion. In Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009) (pp. 2027-2030). (Publications Speech Processing Group, Brno University of Technology). International Speech Communication Association.

Truong, Khiet Phuong ; van Leeuwen, David A. ; Neerincx, Mark A. et al. / Arousal and Valence Prediction in Spontaneous Emotional Speech : Felt versus Perceived Emotion. Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009). International Speech Communication Association, 2009. pp. 2027-2030 (Publications Speech Processing Group, Brno University of Technology).

@inproceedings{40c17ef06e80419c9727933f7734e506,

title = "Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion",

abstract = "In this paper, we describe emotion recognition experiments carried out for spontaneous aﬀective speech with the aim to compare the added value of annotation of felt emotion versus annotation of perceived emotion. Using speech material available in the TNO-GAMING corpus (a corpus containing audiovisual recordings of people playing videogames), speech-based aﬀect recognizers were developed that can predict Arousal and Valence scalar values. Two types of recognizers were developed in parallel: one trained with felt emotion annotations (generated by the gamers themselves) and one trained with perceived/observed emotion annotations (generated by a group of observers). The experiments showed that, in speech, with the methods and features currently used, observed emotions are easier to predict than felt emotions. The results suggest that recognition performance strongly depends on how and by whom the emotion annotations are carried out.",

keywords = "IR-68948, EWI-17024, METIS-264250",

author = "Truong, {Khiet Phuong} and {van Leeuwen}, {David A.} and Neerincx, {Mark A.} and {de Jong}, {Franciska M.G.}",

year = "2009",

language = "English",

series = "Publications Speech Processing Group, Brno University of Technology",

publisher = "International Speech Communication Association",

pages = "2027--2030",

booktitle = "Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009)",

address = "France",

note = "10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009, INTERSPEECH ; Conference date: 06-09-2009 Through 10-09-2009",

url = "http://www.interspeech2010.jpn.org/",

}

Truong, KP, van Leeuwen, DA, Neerincx, MA & de Jong, FMG 2009, Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion. in Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009). Publications Speech Processing Group, Brno University of Technology, International Speech Communication Association, pp. 2027-2030, 10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009, Brighton, United Kingdom, 6/09/09.

Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion. / Truong, Khiet Phuong; van Leeuwen, David A.; Neerincx, Mark A. et al.
Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009). International Speech Communication Association, 2009. p. 2027-2030 (Publications Speech Processing Group, Brno University of Technology).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Arousal and Valence Prediction in Spontaneous Emotional Speech

T2 - 10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009

AU - Truong, Khiet Phuong

AU - van Leeuwen, David A.

AU - Neerincx, Mark A.

AU - de Jong, Franciska M.G.

N1 - Conference code: 10

PY - 2009

Y1 - 2009

N2 - In this paper, we describe emotion recognition experiments carried out for spontaneous aﬀective speech with the aim to compare the added value of annotation of felt emotion versus annotation of perceived emotion. Using speech material available in the TNO-GAMING corpus (a corpus containing audiovisual recordings of people playing videogames), speech-based aﬀect recognizers were developed that can predict Arousal and Valence scalar values. Two types of recognizers were developed in parallel: one trained with felt emotion annotations (generated by the gamers themselves) and one trained with perceived/observed emotion annotations (generated by a group of observers). The experiments showed that, in speech, with the methods and features currently used, observed emotions are easier to predict than felt emotions. The results suggest that recognition performance strongly depends on how and by whom the emotion annotations are carried out.

AB - In this paper, we describe emotion recognition experiments carried out for spontaneous aﬀective speech with the aim to compare the added value of annotation of felt emotion versus annotation of perceived emotion. Using speech material available in the TNO-GAMING corpus (a corpus containing audiovisual recordings of people playing videogames), speech-based aﬀect recognizers were developed that can predict Arousal and Valence scalar values. Two types of recognizers were developed in parallel: one trained with felt emotion annotations (generated by the gamers themselves) and one trained with perceived/observed emotion annotations (generated by a group of observers). The experiments showed that, in speech, with the methods and features currently used, observed emotions are easier to predict than felt emotions. The results suggest that recognition performance strongly depends on how and by whom the emotion annotations are carried out.

KW - IR-68948

KW - EWI-17024

KW - METIS-264250

M3 - Conference contribution

T3 - Publications Speech Processing Group, Brno University of Technology

SP - 2027

EP - 2030

BT - Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009)

PB - International Speech Communication Association

Y2 - 6 September 2009 through 10 September 2009

ER -

Truong KP, van Leeuwen DA, Neerincx MA, de Jong FMG. Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion. In Proceedings of the 10th Annual Conference of the International Speech Communication Association (Interspeech 2009). International Speech Communication Association. 2009. p. 2027-2030. (Publications Speech Processing Group, Brno University of Technology).

Arousal and Valence Prediction in Spontaneous Emotional Speech: Felt versus Perceived Emotion

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this