Two approaches

In its pursuit of combatcapable voicerecognition applications, researchers have taken two approaches.

In its pursuit of combat-capable voice-recognition applications, researchers have taken two approaches.

The first is the use of technology that filters out background noise, such as helicopter rotors and squawking radios.

But filtering can only go so far, so another approach adds to the noise, in a sense. "With the level of noise we're talking about, there is no amount of filtering that can get the noise out," said Mark Bradley, program manager for tactical noise voice-recognition systems at ITT Industries' Aerospace/ Communications Division.

The ITT approach adds noise to a template of speech that would be used in various tactical conditions. When the appropriate template is used, it enables the system to allow for environmental sounds — the whine of a tank engine or the din of conversation in a command center — while recognizing voice commands.

NEXT STORY: Payment due despite NMCI delays