Context-aware signal conditioning for vehicle exterior voice assistant
AssigneeGM Global Technology Operations LLC
InventorsAlaa M. Khamis, Gaurav Talwar, Romeo D. Garcia, Jr., Carmine F. D'agostino, Neeraj R. Gautama
Life of this patent
Estimating the term…
Abstract
A vehicle includes a plurality of microphones to obtain speech from a person outside the vehicle as an input signal and a sensor system to determine a location and orientation of the person relative to the vehicle. The vehicle also includes a controller to determine characteristics of the input signal and to determine whether to perform speech enhancement on the input signal based on one or more of the characteristics and the location and orientation of the person.
Claims20
A vehicle comprising:
- a plurality of microphones configured to obtain speech from a person outside the vehicle as an input signal;
- a sensor system configured to determine a location and an orientation of the person relative to the vehicle, wherein the orientation of the person is determined using a volume level measured at each of the plurality of microphones, wherein the volume level is calculated bsed on a root mean square sound pressure and a reference sound pressure;
- and
- a controller configured to determine characteristics of the input signal and to determine whether to perform speech enhancement on the input signal based on one or more of the characteristics and the location and the orientation of the person.
A method in a vehicle, comprising:
- arranging a plurality of microphones to obtain speech from a person outside the vehicle as an input signal;
- arranging a sensor system to determine a location and an orientation of the person relative to the vehicle;
- and
- configuring a controller to determine characteristics of the input signal and to determine whether to perform speech enhancement on the input signal based on one or more of the characteristics and the location and the orientation of the person,
- wherein the characteristics of the input signal comprise a signal-to-noise ration (SNR), a source-to-artefacts ratio (SAR), a source-to-distortion ration (SDR), and a source-to-interference ration (SIR).
Dependencies are inferred from the claim text.