An Evaluation of a Spoken Language Interface

NOTICES Disclaimers The findings in this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. Citation of manufacturer's or trade names does not constitute an official endorsement or approval of the use thereof. DESTRUCTION NOTICEDestroy this report when it is no longer needed. Do not return it to the originator. Approved for public release; distribution is unlimited. Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden, Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. Speech recognition software has been proposed for inclusion in the Future Force Warrior system. DynaSpeak is a speech recognition software product developed by SRI International that has been shown to function well in quiet environments. However, the system has not yet been evaluated in noise environments representative of Army operational environments. This report summarizes a study that was conducted to evaluate the performance of the DynaSpeak software in the presence of two types of noise: steady state noise obtained from a tracked vehicle and impulse noise obtained from gunfire. Simultaneous recordings were made from 12 participants wearing two microphones: a Gentex noise-canceling boom microphone and a Temco HG-17 bone conduction microphone. The 12 participants verbalized call signs and commands which were analyzed by the speech recognition software. The results of the present study show that the DynaSpeak software functions well through the boom microphone in moderately high steady state noise (90 decibels A-weighted [dBA]) with error rates for words of 2% to 5%. However, the performance of the system with the same boom microphone in high steady state noise levels (110 dBA) or in moderately high impulse noise levels (90 dBA) does not meet the requirements for military applications. Error rates for words in these conditions ranged from 14% to 40%. If the DynaSpeak software is to be used in environments of impulse noise or high levels of steady state noise, the …