Augmented hearing via adaptive self-reinforcement
An audio system receives sounds from a local area with a set of sensors. The audio system analyzes the sounds to identify an amount of background noise in the environment. The audio system may also identify target sound in the environment, such as voice from a nearby speaker, to reinforce. The amount of background noise is then used to adaptively set a level of reinforcement for the target sound. In low-noise environments, the target sound may be minimally, or not reinforced, while in high-noise environments a reinforcement sound signal may be generated with the target sound so that a transducer array may generate reinforcing sounds for the target sound and increase its contrast with the noise in the environment. By applying the reinforcement dynamically, the system automatically adjusts for different sounds and environments.
1. An audio system comprising:
a sensor array configured to capture sound from a local area;
a controller configured to:
localize each particular sound, of multiple sounds in the captured sound, based on:
identified values, for a given sound feature, as differently measured by at least two sensors in the sensor array, and/or
visual information, received from a headset associated with the audio system, indicating location changes for the particular sound,
identify, based at least in part on the localization for the multiple sounds,
A) someone speaking as a sound source in the captured sound and
B) background noise,
determine an amount of the background noise in the captured sound,
determine a level of reinforcement to be applied to at least the sound source based in part on the amount of the background noise, and
generate a reinforcement audio signal based on the level of reinforcement; and
a transducer array configured to generate reinforcement audio for the sound from the sound source based on the reinforcement audio signal.
2. The audio system of claim 1 , wherein the level of reinforcement increases when the amount of background noise increases.
3. The audio system of claim 1 , wherein the controller is further configured to isolate sound from the sound source and the reinforcement audio is based on the isolated sound from the sound source.
4. The audio system of claim 1 , wherein the localizing each particular sound includes creating a virtual model of a local area in which the audio system and the multiple sounds are mapped.
5. The audio system of claim 1 ,
wherein the localizing at least one of the multiple sounds is based on the identified values, for the given sound feature, as differently measured by the at least two sensors in the sensor array, and
wherein the given sound feature is one of: intensity, spectra, and/or arrival time.
6. The audio system of claim 1 , wherein the controller is further configured to identify sound from the sound source based on a language of the sound source.
7. The audio system of claim 1 , wherein the localizing at least one of the multiple sounds is based on the visual information, received from the headset associated with the audio system, indicating location changes for the particular sound.
8. The audio system of claim 1 , wherein the localizing at least one of the multiple sounds is based on the identified values, for the given sound feature, as differently measured by the at least two sensors in the sensor array.
9. The audio system of claim 1 , wherein the controller is further configured to determine an environment type of the local area based on the captured sound and the level of reinforcement is further determined based on the environment type.
10. The audio system of claim 1 , wherein the level of reinforcement is further based on a signal-to-noise ratio of the sound from the sound source relative to the amount of background noise.
11. A method comprising:
localizing each particular sound, of multiple sounds in captured sound from a sensor array, based on:
identified values, for a given sound feature, as differently measured by at least two sensors in the sensor array, and/or
visual information, received from a headset, indicating location changes for the particular sound;
identifying, based at least in part on the localization for the multiple sounds, A) a sound source in the captured sound and B) background noise,
determining an amount of the background noise in the captured sound;
determining a level of reinforcement to be applied to at least the sound source based in part on the amount of the background noise;
generating a reinforcement audio signal based on the level of reinforcement; and
sending the reinforcement audio signal to a transducer array for generation of a reinforcement audio for the sound from the sound source.
12. The method of claim 11 , wherein the level of reinforcement increases when the amount of background noise increases.
13. The method of claim 11 , further comprising isolating sound from the sound source and the reinforcement audio is based on the isolated sound from the sound source.
14. The method of claim 11 , wherein the localizing each particular sound includes creating a virtual model of a local area in which the audio system and the multiple sounds are mapped.
15. The method of claim 11 ,
wherein the localizing at least one of the multiple sounds is based on the identified values, for the given sound feature, as differently measured by the at least two sensors in the sensor array, and
wherein the given sound feature is one of: intensity, spectra, and/or arrival time.
16. The method of claim 11 , further comprising identifying sound from the sound source based on a language of the sound source.
17. The method of claim 11 , wherein the localizing at least one of the multiple sounds is based on the visual information, received from the headset associated with the audio system, indicating location changes for the particular sound.
18. The method of claim 11 , wherein the localizing at least one of the multiple sounds is based on the identified values, for the given sound feature, as differently measured by the at least two sensors in the sensor array.
19. The method of claim 11 , further comprising determining an environment type based on the captured sound and the level of reinforcement is further determined based on the environment type.
20. The method of claim 11 , wherein the level of reinforcement is further based on a signal-to-noise ratio of the sound from the sound source relative to the amount of background noise.