Detail publikace

TRAP based features for LVCSR of meeting data

KARAFIÁT, M., GRÉZL, F., ČERNOCKÝ, J.

Originální název

TRAP based features for LVCSR of meeting data

Anglický název

TRAP based features for LVCSR of meeting data

Jazyk

en

Originální abstrakt

This paper describes using temporal patterns (TRAPs) feature extraction in large vocabulary continuous speech recognition (LVCSR) of meeting data. Frequency differentiation and local operators are applied to critical-band speech spectrum. Tests are performed with HMM recognizer on ICSI meetings database. We show that TRAP features in combination with standard ones lead to improvement of word-error rate (WER).

Anglický abstrakt

This paper describes using temporal patterns (TRAPs) feature extraction in large vocabulary continuous speech recognition (LVCSR) of meeting data. Frequency differentiation and local operators are applied to critical-band speech spectrum. Tests are performed with HMM recognizer on ICSI meetings database. We show that TRAP features in combination with standard ones lead to improvement of word-error rate (WER).

Dokumenty

BibTex


@inproceedings{BUT17131,
  author="Martin {Karafiát} and František {Grézl} and Jan {Černocký}",
  title="TRAP based features for LVCSR of meeting data",
  annote="This paper describes using temporal patterns (TRAPs) feature
  extraction in large vocabulary continuous speech recognition (LVCSR) of
  meeting data. Frequency differentiation and local operators are
  applied to critical-band speech spectrum. 
  Tests are performed with HMM recognizer on ICSI meetings
  database.
  We show that TRAP features in combination with standard ones lead to
  improvement of word-error rate (WER).",
  address="Sunjin Printing Co,",
  booktitle="Proc. 8th International Conference on Spoken Language Processing",
  chapter="17131",
  institution="Sunjin Printing Co,",
  journal="8th International Conference on Spoken Language Processing",
  number="10",
  year="2004",
  month="june",
  pages="437--440",
  publisher="Sunjin Printing Co,",
  type="conference paper"
}