Automatic Speaker Recognition for Military Applications: Applications Survey and Operational Requirements

摘要

Automatic speaker recognition (ASR) systems, capable of identifying a person based on voice input alone, have a wide range of potential applications in Navy environments. They could provide access control to restricted areas, equipment and information. They could verify the identity of users of various communication channels, or verify computer users through terminals accepting voice input. This report presents the results of a survey conducted to determine which type of application offers the best potential use of ASR for the Navy. The survey results also give a rough indication of the operational requirements of ASR systems in Naval environments. In designing an ASR system for a particular application there are a number of factors regarding the user and the environment that must be taken into consideration. These factors, and their impact on the operational requirements, are discussed in detail. The appendix of this report outlines the various approaches to ASR and the major ASR systems described in the literature. (Author)