Background noise picked up by microphones can be mitigated using audio beamforming. However, if a user loses alignment with the beam, the user’s speech can be severely attenuated or lost in the noise. This disclosure utilizes a camera, with user permission, to obtain coordinate information for a user that is providing audio input, e.g., in a voice/video call, during an audio recording, etc. The information is utilized to dynamically adjust the direction of the beam of an audio beamformer such that a user’s speech, e.g., as received by the beamformer is consistent regardless of the current position or movement of the user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Igo, Chris, "Camera-driven Beamforming for Better Audio Capture", Technical Disclosure Commons, (May 20, 2021)