Pregled bibliografske jedinice broj: 292414
Efficient Zero-Crossing-Based Autoregressive Feature Extraction for Speech Recognition
Efficient Zero-Crossing-Based Autoregressive Feature Extraction for Speech Recognition // Alta Frequenza, 57 (1988), 8; 501-504 (međunarodna recenzija, članak, znanstveni)
CROSBI ID: 292414 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Efficient Zero-Crossing-Based Autoregressive Feature Extraction for Speech Recognition
Autori
Lipovac, Vladimir
Izvornik
Alta Frequenza (0002-6557) 57
(1988), 8;
501-504
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
autoregressive; speech; zero-crossing
Sažetak
In this paper, a modification of the linear- predictive-coding (LPC) signal analysis is presented, where the polar-coincidence estimation of the autocorrelation function was applied, enabling significant reduction of the data coding rate with certain signal classes. Applying this to the speech signal analysis enables very good results at both the level of the signal coding quality and the accuracy of the single word recognition.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika
Citiraj ovu publikaciju:
Časopis indeksira:
- Current Contents Connect (CCC)
- Web of Science Core Collection (WoSCC)
- SCI-EXP, SSCI i/ili A&HCI
- Scopus
