EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Speech Emotion Recognition Incorporating Relative Difficulty and Labeling Reliability

Youngdo Ahn iD, Sangwook Han iD, Seonggyu Lee iD, Jong Won Shin iD

Sensors · 2024

DOI: 10.3390/s24134111 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, 2024. Verfasst von Youngdo Ahn und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology, Buk-gu, Gwangju 61005, Republic of Korea. Hauptthemen laut MEDLINE: Emotions, Speech. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Korea government(MSIT), Institute of Information and communications Technology Planning and Evaluation (IITP), Korean government (MSIT).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Emotions in speech are expressed in various ways, and the speech emotion recognition (SER) model may perform poorly on unseen corpora that contain different emotional factors from those expressed in training databases. To construct an SER model robust to unseen corpora, regularization approaches or metric losses have been studied. In this paper, we propose an SER method that incorporates relative difficulty and labeling reliability of each training sample. Inspired by the Proxy-Anchor loss, we propose a novel loss function which gives higher gradients to the samples for which the emotion labels are more difficult to estimate among those in the given minibatch. Since the annotators may label the emotion based on the emotional expression which resides in the conversational context or other modality but is not apparent in the given speech utterance, some of the emotional labels may not be reliable and these unreliable labels may affect the proposed loss function more severely. In this regard, we propose to apply label smoothing for the samples misclassified by a pre-trained SER model. Experimental results showed that the performance of the SER on unseen corpora was improved by adopting the proposed loss function with label smoothing on the misclassified data.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

EmotionsSpeechHumansAlgorithmsReproducibility of ResultsPattern Recognition, AutomatedDatabases, Factual

Stichwörter der Autor:innen

generalizationlabeling reliabilityout-of-corpusrelative difficultyspeech emotion recognition

Autor:innen und Institutionen

  • School of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology, Buk-gu, Gwangju 61005, Republic of Korea.

ORCID-Profile: Youngdo Ahn, Sangwook Han, Seonggyu Lee, Jong Won Shin

Förderung

  • Korea government(MSIT)
  • Institute of Information and communications Technology Planning and Evaluation (IITP)
  • Korean government (MSIT)

Zitieren

APA

Ahn, Y., Han, S., Lee, S., & Shin, J. W. (2024). Speech Emotion Recognition Incorporating Relative Difficulty and Labeling Reliability. Sensors. https://doi.org/10.3390/s24134111

Vancouver

Ahn Y, Han S, Lee S, Shin JW. Speech Emotion Recognition Incorporating Relative Difficulty and Labeling Reliability. Sensors. 2024. doi:10.3390/s24134111

BibTeX
@article{Ahn2024,
  author = {Ahn, Youngdo and Han, Sangwook and Lee, Seonggyu and Shin, Jong Won},
  title = {Speech Emotion Recognition Incorporating Relative Difficulty and Labeling Reliability},
  journal = {Sensors},
  year = {2024},
  doi = {10.3390/s24134111}
}
Kostenlos starten →