Voice response systems utilizing digital voice storage have existed for several years. The advantages of using low bit rate speech coding methods in these applications are also well known (1,2). However, few commercial voice response systems to date have performed any compression of the digitized speech signal. This has been due to the fact that toll- or near-toll-quality speech is generally required; until recently, this was only possible with relatively high bit rate waveform coders. The paper will describe the implementation of and our experiences with real-time mutlichannel Code-Excited Linear Predictive (CELP) speech coding multiple digital signal processors.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.