Voice recognition in a loud environment
Hear the difference audio processing makes to voice recognition
Voice recognition is an emerging technology for touchless control of a self-service device. Jointly developed by AWS and Elenium for loud environments, our patent pending approach to voice recognition distinguishes voices in busy environments like airports, hospitals, or medical centres by combining multiple types of sensors, including proximity sensors, cameras and multiple directional microphones.
When a person is detected within the proximity of the device the camera system starts to track the person’s lips. The system then localises which microphone in the array best aligns with the person’s voice. That microphone becomes the primary audio source, while remaining microphones in the array are used to capture background noise. The background noise is subtracted from the final output transmitted to text services by AWS.
Read more about Touchless self-service and vital sign detection