| 摘要 |
Voice control systems are used in diverse technical fields. In this case, the spoken words are firstly detected, usually by one or more microphones (M1, M2) and then fed to a speech recognition system (REC), which determines from the detected signals those words having the highest probability. The robustness with respect to acoustic interfering influences can be increased by the microphone being fastened or held directly in front of the speaker's mouth. The invention enables voice control even from a relatively large distance by virtue of the fact that interfering background noises are spatially separated by the use of directional microphones or microphone arrays (M1, M2). In this case, the directional characteristic is adapted by virtue of the fact that means for detecting the whereabouts of the user (IR1, VR, CIR) are provided. One or more infrared motion detectors are preferably used for this purpose. <IMAGE> |
