Publication detail

Method for Recognition the Actual Emotional State of Speaker

STANĚK, M.

Original Title

Method for Recognition the Actual Emotional State of Speaker

Type

conference paper

Language

English

Original Abstract

This paper is focused on the differences of important points in LPC spectrum depending on emotional state for Czech vowels. LPC spectrum contains antiformats- the local minima, for-mants- the local peaks and its points of inflection. The position of all important points depends on actual emotional state of speaker. Significant differences in type of distribution were obtained for each important point of LPC spectrum within the emotional states of speaker database for F1 F2 and F2-F3 band. Most similar shapes of histograms were achieved for alcohol intoxication and happiness. Ordinary values of relevant histogram parameters for each important point within the emotional states of speaker database are a matter of next statistical investigation.

Keywords

speech processing, emotion recognition, formant analysis

Authors

STANĚK, M.

RIV year

2013

Released

26. 4. 2013

Location

Brno

ISBN

978-80-214-4695-3

Book

Proceedings of the 19th Conference Student EEICT 2013 Volume3

Edition number

1

Pages from

55

Pages to

59

Pages count

5

BibTex

@inproceedings{BUT99256,
  author="Miroslav {Staněk}",
  title="Method for Recognition the Actual Emotional State of Speaker",
  booktitle="Proceedings of the 19th Conference Student EEICT 2013 Volume3",
  year="2013",
  number="1",
  pages="55--59",
  address="Brno",
  isbn="978-80-214-4695-3"
}