Detail publikace

Long Time Analysis of Glottal Pulse Waveform in Stressed Speech Recognition

BOŠTÍK, M., SIGMUND, M.

Originální název

Long Time Analysis of Glottal Pulse Waveform in Stressed Speech Recognition

Anglický název

Long Time Analysis of Glottal Pulse Waveform in Stressed Speech Recognition

Jazyk

en

Originální abstrakt

In the paper, a new technique of speech pro-cessing in emotional stress recognition framework is descri-bed. Features such as pitch, intensity or formant frequencies are usually used for detecting emotional stress in speech. However, in this work, glottal pulses were extracted from speech signal by iterative adaptive inverse filtering technique and a new graphical evaluation technique based on fixed pulse waveform was used. Speech data were collected during oral final examinations supposing students were under the examination stress.

Anglický abstrakt

In the paper, a new technique of speech pro-cessing in emotional stress recognition framework is descri-bed. Features such as pitch, intensity or formant frequencies are usually used for detecting emotional stress in speech. However, in this work, glottal pulses were extracted from speech signal by iterative adaptive inverse filtering technique and a new graphical evaluation technique based on fixed pulse waveform was used. Speech data were collected during oral final examinations supposing students were under the examination stress.

Dokumenty

BibTex 


@inproceedings{BUT11214,
  author="Milan {Boštík} and Milan {Sigmund}",
  title="Long Time Analysis of Glottal Pulse Waveform in Stressed Speech Recognition",
  annote="In the paper, a new technique of speech pro-cessing in emotional stress recognition framework is descri-bed. Features such as pitch, intensity or formant frequencies are usually used for detecting emotional stress in speech. However, in this work, glottal pulses were extracted from speech signal by iterative adaptive inverse filtering technique and a new graphical evaluation technique based on fixed pulse waveform was used. Speech data were collected during oral final examinations supposing students were under the examination stress.",
  address="PTETiS",
  booktitle="Proceedings of IWSSIP 2004",
  chapter="11214",
  institution="PTETiS",
  year="2004",
  month="january",
  pages="311",
  publisher="PTETiS",
  type="conference paper"
}