Autonomous vehicle operation using acoustic modalities
Techniques for autonomous vehicle operation using acoustic modalities include using one or more acoustic sensors of a vehicle to receive acoustic waves from one or more objects. The acoustic waves have multiple wavelengths. The acoustic waves are clustered into one or more acoustic clusters based on the plurality of wavelengths. A particular acoustic cluster of the one or more acoustic clusters is selected based on signal processing of the one or more acoustic clusters. A particular object is associated with the particular acoustic cluster. An acoustic fingerprint of the particular object is generated based on the particular acoustic cluster. Characteristics of the particular object are determined based on the acoustic fingerprint of the particular object. A control circuit of the vehicle is used to operate the vehicle to avoid a collision with the particular object based on the characteristics of the particular object.
1. A method comprising:
receiving, using one or more acoustic sensors of a vehicle, acoustic waves from two or more objects, wherein the acoustic waves have a plurality of wavelengths;
clustering, using one or more processors of the vehicle, the acoustic waves into two or more acoustic clusters based on the plurality of wavelengths;
selecting, using the one or more processors, a particular acoustic cluster of the two or more acoustic clusters based on signal processing of the two or more acoustic clusters;
associating, using the one or more processors, a particular object of the two or more objects with the particular acoustic cluster;
generating, using the one or more processors, an acoustic fingerprint of the particular object based on the particular acoustic cluster;
determining, using the one or more processors, characteristics of the particular object based on the acoustic fingerprint of the particular object; and
operating, using a control circuit of the vehicle, the vehicle to avoid a collision with the particular object based on the characteristics of the particular object.
2. The method of claim 1 , wherein each acoustic cluster of the ene two or more acoustic clusters corresponds to a respective object of the one two or more objects.
3. The method of claim 1 , wherein the clustering of the acoustic waves comprises determining, using the one or more processors, a change in a wavelength or a frequency of the acoustic waves based on filtering the acoustic waves.
4. The method of claim 3 , wherein the filtering of the acoustic waves is performed using a gradient filter.
5. The method of claim 1 , wherein the clustering of the acoustic waves comprises splitting, using the one or more processors, the acoustic waves into the two or more acoustic clusters based on a spectrogram.
6. The method of claim 1 , wherein the clustering of the acoustic waves comprises splitting, using the one or more processors, the acoustic waves into the two or more acoustic clusters based on a time-frequency graph.
7. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, a zero crossing rate of the particular acoustic cluster of the two or more acoustic clusters.
8. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, a spectral density of the particular acoustic cluster of the two or more acoustic clusters.
9. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, a spectral flatness of the particular acoustic cluster of the two or more acoustic clusters.
10. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, peaks in a spectrogram of the particular acoustic cluster of the two or more acoustic clusters.
11. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, an acoustic pressure of the particular acoustic cluster of the two or more acoustic clusters.
12. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, a particle velocity of the particular acoustic cluster of the two or more acoustic clusters.
13. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, a particle displacement of the particular acoustic cluster of the two or more acoustic clusters.
14. The method of claim 1 , wherein the generating of the acoustic fingerprint comprises determining, using the one or more processors, an acoustic intensity of the particular acoustic cluster of the two or more acoustic clusters.
15. The method of claim 1 , wherein the acoustic waves are diffracted, reflected, or interfered with by the particular object of the two or more objects.
16. The method of claim 1 , further comprising filtering, using the one or more processors, the acoustic waves based on peaks in a spectrogram of the acoustic waves to remove background noise.
17. The method of claim 1 , wherein the characteristics comprise physical characteristics of the particular object, and wherein the determining of the physical characteristics comprises generating, using the one or more processors, a hash of peaks in a spectrogram of the acoustic waves.
18. The method of claim 17 , wherein the determining of the physical characteristics further comprises searching, using the one or more processors, a database based on the hash of peaks to determine the physical characteristics.
19. A vehicle comprising:
one or more computer processors; and
one or more non-transitory storage media storing instructions which, when executed by the one or more computer processors, cause the one or more computer processors to:
receive, using one or more acoustic sensors of a vehicle, acoustic waves from two or more objects, wherein the acoustic waves have a plurality of wavelengths;
cluster the acoustic waves into two or more acoustic clusters based on the plurality of wavelengths;
select a particular acoustic cluster of the two or more acoustic clusters based on signal processing of the two or more acoustic clusters;
associate a particular object of the two or more objects with the particular acoustic cluster;
generate an acoustic fingerprint of the particular object based on the particular acoustic cluster;
determine characteristics of the particular object based on the acoustic fingerprint of the particular object; and
operate, using a control circuit of the vehicle, the vehicle to avoid a collision with the particular object based on the characteristics of the particular object.
20. One or more non-transitory storage media storing instructions which, when executed by one or more computing devices, cause the one or more computing devices to:
receive, using one or more acoustic sensors of a vehicle, acoustic waves from two or more objects, wherein the acoustic waves have a plurality of wavelengths;
cluster the acoustic waves into two or more acoustic clusters based on the plurality of wavelengths;
select a particular acoustic cluster of the two or more acoustic clusters based on signal processing of the two or more acoustic clusters;
associate a particular object of the two or more objects with the particular acoustic cluster;
generate an acoustic fingerprint of the particular object based on the particular acoustic cluster;
determine characteristics of the particular object based on the acoustic fingerprint of the particular object; and
operate, using a control circuit of the vehicle, the vehicle to avoid a collision with the particular object based on the characteristics of the particular object.