Third European Conference on Speech Communication and Technology

Berlin, Germany
September 22-25, 1993


A Baseline of a Speaker Independent Continuous Speech Recognizer of Italian

B. Angelini, F. Brugnara, D. Falavigna, D. Giuliani, R. Gretter, M. Omologo

IRST-Istituto per la Ricerca Sdentifica e Tecnologica, Povo di Trento, Italy

The objective of this paper is to describe some preliminary results of the activity that is being carried out in IRST laboratories for the development of an HMM-based speaker independent continuous speech recognition system for the Italian language. The recognition system is being developed using the acoustic-phonetic continuous speech portion of APASCI database, also described in this paper. Performance is reported for different recognition tasks that were considered: without any phonetic constraint, the system provided a 72% Phone Accuracy; further, given a vocabulary size of 1019 words and no lexical constraints, a 73% Word Accuracy was achieved.

Keywords: Speaker Independent Continuous Speech Recognition, Speech Databases, Hidden Markov Models.

Full Paper

Bibliographic reference.  Angelini, B. / Brugnara, F. / Falavigna, D. / Giuliani, D. / Gretter, R. / Omologo, M. (1993): "A baseline of a speaker independent continuous speech recognizer of Italian", In EUROSPEECH'93, 847-850.