DocumentCode
3245469
Title
Progress in deployment and further development of the NYNEX VoiceDialingSM service
Author
Vysotsky, George J.
Author_Institution
NYNEX Sci. & Technol. Inc., White Plains, NY, USA
fYear
1996
fDate
30 Sep-1 Oct 1996
Firstpage
16
Lastpage
20
Abstract
As was reported at IVTTA-94, the NYNEX VoiceDialingSM service has been deployed, starting from the middle of 1993, in the NYNEX as well as other Bell Operating Companies regions. Since 1995, several substantial developments have taken place. One of them is the transition from a DTW based recognition algorithm implemented in hardware to a general purpose DSP implementation of a continuous density multi-gaussian mixture HMM based algorithm. As a result, it was possible to expand the service beyond speaker dependent name recognition. The paper highlights major efforts in this transition and provides a more detailed description of certain system components
Keywords
Gaussian processes; digital signal processing chips; hidden Markov models; speech recognition; telephony; Bell Operating Companies; DSP implementation; DTW; NYNEX VoiceDialing service; continuous density multi-gaussian mixture HMM; hidden Markov model; performance; speaker dependent name recognition; speech recognition algorithm; telephony; Cepstral analysis; Covariance matrix; Digital signal processing; Hardware; Hidden Markov models; Prototypes; Speech enhancement; Speech processing; Speech recognition; Viterbi algorithm;
fLanguage
English
Publisher
ieee
Conference_Titel
Interactive Voice Technology for Telecommunications Applications, 1996. Proceedings., Third IEEE Workshop on
Conference_Location
Basking Ridge, NJ
Print_ISBN
0-7803-3238-5
Type
conf
DOI
10.1109/IVTTA.1996.552710
Filename
552710
Link To Document