Robust speech recognition techniques evaluation for telephony server based in-car applications

• Published in: 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing Vol. 1; pp. I - 65
• Main Author: Delphin-Poulat, L.
• Format: Conference Proceeding
• Language: English
• Published: Piscataway, N.J IEEE 2004
• Subjects: Applied sciences; Equipments and installations; Exact sciences and technology; Feature extraction; Information, signal and communications theory; Land mobile radio cellular systems; Microphone arrays; Mobile radiocommunication systems; Network servers; Noise robustness; Performance gain; Performance loss; Propagation losses; Radiocommunications; Signal processing; Speech processing; Speech recognition; Systems, networks and services of telecommunications; Telecommunications; Telecommunications and information theory; Telephony; Transmission and modulation (techniques and equipments); Performance evaluation; Mobile radiocommunication; Sound recording; Microphone; GSM system; Acoustic wave; Wireless telecommunication; Man machine dialogue; Cell network; Wave transmission; Cepstral analysis; Audio systems; User interface; Sound transmission; Speech recognition; Telecommunication network; Signal processing; Telephony; Feature extraction; Noise immunity; Feasibility; Sensor array; Speech processing
• ISBN: 9780780384842; 0780384849
• ISSN: 1520-6149; 2379-190X
• DOI: 10.1109/ICASSP.2004.1325923

In this paper, the feasibility of designing a speech-recognition based telephony server for in-car applications with an acceptable recognition rate is investigated. The whole acoustic channel (sound pickup, sound transmission over the cellular network, feature extraction) is evaluated: the loss or the gain in performance due to each element is quantified. More precisely, two sound pickup systems (a hypercardioid microphone and a microphone array) were tested. A standard MFCC and the Aurora advanced front-ends were evaluated. Recognition performance was measured before and after transmission over a cellular (GSM) network. The gain of using either a robust sound recording device or noise robust front-end is demonstrated.