Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

ASR for emotional speech: clarifying the issues and enhancing performance.

T Athanaselis1, S Bakamidis, I Dologlou

  • 1Department of Speech Technology, Institute for Language and Speech Processing (ILSP), Artemidos 6 & Epidavrou, GR-151 25 Maroussi, Athens, Greece. tathana@ilsp.gr

Neural Networks : the Official Journal of the International Neural Network Society
|June 11, 2005
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Encephalopathy associated with respiratory syncytial virus bronchiolitis.

Journal of child neurology·2001
Same author

Incidental discovery of pelvic cholelithiasis on diagnostic laparoscopy.

Surgical endoscopy·2001
Same author

Methylmercury and neurodevelopment: reanalysis of the Seychelles Child Development Study outcomes at 66 months of age.

JAMA·2001
Same author

Depression and self-reported functional status in older primary care patients.

The American journal of psychiatry·2001
Same author

A randomized, controlled trial of remacemide for motor fluctuations in Parkinson's disease.

Neurology·2001
Same author

Acute pulmonary effects of ultrafine particles in rats and mice.

Research report (Health Effects Institute)·2001

Recognizing emotional speech is challenging. Enhancing language models with more emotional content significantly improves the recognition of spontaneous emotional speech by approximately 20%.

Area of Science:

  • Speech processing
  • Computational linguistics
  • Affective computing

Background:

  • Recognizing verbal content in emotional speech is a known challenge with low reported success rates.
  • Prosody aids recognition in acted emotions, but its effectiveness in spontaneous speech remains unproven.

Purpose of the Study:

  • To investigate methods for improving the recognition rate of spontaneous emotional speech.
  • To determine if language models incorporating more emotional data can enhance speech emotion recognition.

Main Methods:

  • Adapted the British National Corpus (BNC) by incorporating an emotional lexicon to identify and increase the representation of emotional utterances.
  • Developed and evaluated a language model based on this augmented corpus for spontaneous emotional speech recognition.

Related Experiment Videos

Main Results:

  • The language model utilizing increased emotional utterance representation improved recognition rates by approximately 20%.
  • This demonstrates the effectiveness of enhanced language models for spontaneous emotional speech.

Conclusions:

  • Increasing the representation of emotional utterances in language models is a viable strategy to improve spontaneous emotional speech recognition.
  • The findings suggest a practical approach to enhance the performance of speech emotion recognition systems.