Dynamic head-related transfer function

1. A method comprising:

receiving audio recordings of a sound captured by each of a plurality of audio sensors, the sound emitted by an in-ear speaker worn by a subject;

receiving a reference signal captured by a microphone coupled to the in-ear speaker;

receiving one or more images captured by each of a plurality of image sensors, the one or more images depicting a body pose of the subject while the sound is emitted by the in-ear speaker;

generating a pose representation of the body pose of the subject based on the one or more images;

determining a head-related transfer function (HRTF) for each of the plurality of audio sensors based on the pose representation, the audio recordings of the sound, and the reference signal; and

determining a dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors.

2. The method of claim 1 , wherein determining the dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors further comprises:

receiving audio recordings of an additional sound captured by each of the plurality of audio sensors, the additional sound emitted by the in-ear speaker worn by the subject;

receiving an additional reference signal captured by the microphone coupled to the in-ear speaker;

receiving an additional one or more images captured by each of the plurality of image sensors, the additional one or more images depicting an additional body pose of the subject while the additional sound is emitted by the in-ear speaker;

modifying the pose representation of the body pose of the subject based on the additional one or more images; and

determining the HRTF for each of the plurality of audio sensors based on the pose representation, the audio recordings of the additional sound, and the additional reference signal.

3. The method of claim 1 , wherein the sound captured by each of the plurality of audio sensors comprises a sine wave sweep.

4. The method of claim 1 , wherein the body pose of the subject indicates an orientation of the subject in relation to a surrounding capture space.

5. The method of claim 4 , wherein the plurality of audio sensors and the plurality of image sensors are disposed evenly throughout the surrounding capture space.

6. The method of claim 1 , wherein the HRTF for each of the plurality of audio sensors is stored in an HRTF database.

7. The method of claim 1 , wherein determining the HRTF for each of the plurality of audio sensors comprises at least one of:

identifying, for each of the plurality of audio sensors, an azimuth of the subject in relation to the audio sensor;

identifying, for each of the plurality of audio sensors, an elevation of the subject in relation to the audio sensor;

identifying, for each of the plurality of audio sensors, a radius between the subject and the audio sensor; and

identifying, for each of the plurality of audio sensors, the body pose of the subject in relation to the audio sensor.

8. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

receive audio recordings of a sound captured by each of a plurality of audio sensors, the sound emitted by an in-ear speaker worn by a subject;

receive a reference signal captured by a microphone coupled to the in-ear speaker;

receive one or more images captured by each of a plurality of image sensors, the one or more images depicting a body pose of the subject while the sound is emitted by the in-ear speaker;

generate a pose representation of the body pose of the subject based on the one or more images;

determine a head-related transfer function (HRTF) for each of the plurality of audio sensors based on the pose representation, the audio recordings of the sound, and the reference signal; and

determine a dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors.

9. The media of claim 8 , wherein to determine the dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors, the software is further operable when executed to:

receive audio recordings of an additional sound captured by each of the plurality of audio sensors, the additional sound emitted by the in-ear speaker worn by the subject;

receive an additional reference signal captured by the microphone coupled to the in-ear speaker;

receive an additional one or more images captured by each of the plurality of image sensors, the additional one or more images depicting an additional body pose of the subject while the additional sound is emitted by the in-ear speaker;

modify the pose representation of the body pose of the subject based on the additional one or more images; and

determine the HRTF for each of the plurality of audio sensors based on the pose representation, the audio recordings of the additional sound, and the additional reference signal.

10. The media of claim 8 , wherein the sound captured by each of the plurality of audio sensors comprises a sine wave sweep.

11. The media of claim 8 , wherein the body pose of the subject indicates an orientation of the subject in relation to a surrounding capture space.

12. The media of claim 11 , wherein the plurality of audio sensors and the plurality of image sensors are disposed evenly throughout the surrounding capture space.

13. The media of claim 8 , wherein the HRTF for each of the plurality of audio sensors is stored in an HRTF database.

14. The media of claim 8 , wherein to determine the HRTF for each of the plurality of audio sensors, the software is further operable when executed to:

identify, for each of the plurality of audio sensors, an azimuth of the subject in relation to the audio sensor;

identify, for each of the plurality of audio sensors, an elevation of the subject in relation to the audio sensor;

identify, for each of the plurality of audio sensors, a radius between the subject and the audio sensor; and

identify, for each of the plurality of audio sensors, the body pose of the subject in relation to the audio sensor.

15. A system comprising:

an in-ear speaker;

a plurality of audio sensors;

a microphone;

a plurality of image sensors;

one or more processors; and

one or more computer-readable non-transitory storage media coupled to one or more of the processors and comprising instructions operable when executed by one or more of the processors to cause the system to:

receive audio recordings of a sound captured by each of the plurality of audio sensors, the sound emitted by the in-ear speaker worn by a subject;

receive a reference signal captured by the microphone coupled to the in-ear speaker;

receive one or more images captured by each of the plurality of image sensors, the one or more images depicting a body pose of the subject while the sound is emitted by the in-ear speaker;

generate a pose representation of the body pose of the subject based on the one or more images;

determine a head-related transfer function (HRTF) for each of the plurality of audio sensors based on the pose representation, the audio recordings of the sound, and the reference signal; and

determine a dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors.

16. The system of claim 15 , wherein to determine the dynamic HRTF for the subject based on the HRTF for each of the plurality of audio sensors, the instructions are further operable when executed to:

receive audio recordings of an additional sound captured by each of the plurality of audio sensors, the additional sound emitted by the in-ear speaker worn by the subject;

receive an additional reference signal captured by the microphone coupled to the in-ear speaker;

receive an additional one or more images captured by each of the plurality of image sensors, the additional one or more images depicting an additional body pose of the subject while the additional sound is emitted by the in-ear speaker;

modify the pose representation of the body pose of the subject based on the additional one or more images; and

determine the HRTF for each of the plurality of audio sensors based on the pose representation, the audio recordings of the additional sound, and the additional reference signal.

17. The system of claim 15 , wherein the sound captured by each of the plurality of audio sensors comprises a sine wave sweep.

18. The system of claim 15 , wherein the body pose of the subject indicates an orientation of the subject in relation to a surrounding capture space.

19. The system of claim 18 , wherein the plurality of audio sensors and the plurality of image sensors are disposed evenly throughout the surrounding capture space.

20. The system of claim 15 , wherein the HRTF for each of the plurality of audio sensors is stored in an HRTF database.