System and method for acoustic speaker localization
A method, computer program product, and computing system for acoustic speech localization, comprising receiving, via a plurality of microphones, a plurality of audio signals. Modulation properties of the plurality of audio signals may be analyzed. Speech sounds may be localized from the plurality of audio signals based upon, at least in part, the modulation properties of the plurality of audio signals.
1. A computer-implemented method for acoustic speech localization, executed on a computing device, comprising:
receiving, via a plurality of microphones, a plurality of audio signals;
analyzing, at the computing device, modulation properties of the plurality of audio signals; and
localizing speech sounds from the plurality of audio signals based upon, at least in part, the modulation properties of the plurality of audio signals.
2. The computer-implemented method of claim 1 , wherein analyzing the modulation properties of the plurality of audio signals includes:
modeling the plurality of audio signals in the modulation domain as a plurality of carrier signals and a plurality of modulator signals.
3. The computer-implemented method of claim 2 , further comprising:
filtering, via one or more filters, a subset of the plurality of modulator signals of the plurality audio signals.
4. The computer-implemented method of claim 2 , further comprising:
subtracting one or more signals associated with one or more modulator signals from the plurality of modulator signals of the plurality of audio signals.
5. The computer-implemented method of claim 1 , wherein localizing speech sounds from the plurality of audio signals includes:
computing a cross-correlation of the plurality of audio signals received by one or more pairs of microphones of the plurality of microphones.
6. The computer-implemented method of claim 5 , wherein localizing speech sounds from the plurality of audio signals includes:
applying a weight to at least a subset of the plurality of audio signals based upon, at least in part, a signal-to-noise ratio (SNR) of the plurality of audio signals.
7. The computer-implemented method of claim 6 , wherein the weight is applied based upon at least in part a probability of speech sounds being present in at least a portion of the plurality of audio signals.
8. A computer program product residing on a non-transitory computer readable medium having a plurality of instructions stored thereon which, when executed by a processor, cause the processor to perform operations comprising:
receiving, via a plurality of microphones, a plurality of audio signals;
analyzing modulation properties of the plurality of audio signals; and
localizing speech sounds from the plurality of audio signals based upon, at least in part, the modulation properties of the plurality of audio signals.
9. The computer program product of claim 8 , wherein analyzing the modulation properties of the plurality of audio signals includes:
modeling the plurality of audio signals in the modulation domain as a plurality of carrier signals and a plurality of modulator signals.
10. The computer program product of claim 9 , further comprising:
filtering, via one or more filters, a subset of the plurality of modulator signals of the plurality audio signals.
11. The computer program product of claim 9 , further comprising:
subtracting one or more signals associated with one or more modulator signals from the plurality of modulator signals of the plurality of audio signals.
12. The computer program product of claim 8 , wherein localizing speech sounds from the plurality of audio signals includes:
computing a cross-correlation of the plurality of audio signals received by one or more pairs of microphones of the plurality of microphones.
13. The computer program product of claim 12 , wherein localizing speech sounds from the plurality of audio signals includes:
applying a weight mask to at least a subset of the plurality of audio signals based upon, at least in part, a signal-to-noise ratio (SNR) of the plurality of audio signals.
14. The computer program product of claim 13 , wherein the weight mask is applied based upon at least in part a probability of speech sounds being present in at least a portion of the plurality of audio signals.
15. A computing system including a processor and memory configured to perform operations comprising:
receiving, via a plurality of microphones, a plurality of audio signals;
analyzing modulation properties of the plurality of audio signals; and
localizing speech sounds from the plurality of audio signals based upon, at least in part, the modulation properties of the plurality of audio signals.
16. The computing system of claim 15 , wherein analyzing the modulation properties of the plurality of audio signals includes:
modeling the plurality of audio signals in the modulation domain as a plurality of carrier signals and a plurality of modulator signals.
17. The computing system of claim 16 , further comprising:
filtering, via one or more filters, a subset of the plurality of modulator signals of the plurality audio signals.
18. The computing system of claim 16 , further comprising:
subtracting one or more signals associated with one or more modulator signals from the plurality of modulator signals of the plurality of audio signals.
19. The computing system of claim 15 , wherein localizing speech sounds from the plurality of audio signals includes:
computing a cross-correlation of the plurality of audio signals received by one or more pairs of microphones of the plurality of microphones.
20. The computing system of claim 19 , wherein localizing speech sounds from the plurality of audio signals includes:
applying a weight mask to at least a subset of the plurality of audio signals based upon, at least in part, a signal-to-noise ratio (SNR) of the plurality of audio signals.