 |
| To ensure reliability of an ASR system, the acoustic component of the system must be tested. This means using speech from the operator position and considering the acoustic aspects of various operating conditions for background noise in any environment. HEAD acoustics offers one-stop services to create database of calibrated speech and noise recordings that are ready to use in automated speech recognition testing software such as HEAD acoustics VoCAS system, the flexible turnkey solution for evaluating voice recognition systems. |
|
|
| |
 |
|
Proven methodology for preparing and accurately playing back a database of speech utterances mixed with background noise from various operating conditions and environments |
|
|
Recording of calibrated background noises from ASR product use cases |
|
|
Recording of specific speech material including Lombard Effect (“Lombard Speech”) when subjects are exposed to background noise in the driving vehicle |
|
|
Calibrated input to the speech recognition engine can either be done acoustically to the frontside of the hands-free microphone or electrically to the backside of the hands-free microphone |
|
|
Testing and following tuning of voice recognition systems |
|
|
|
|
|
|
|
|
|
|
|
Automatic Speech Recognition (ASR)
