Self-localization for a set of microphones
View Patent ↗Provided are methods and systems for finding the location of sensors (e.g., microphones) with unknown internal delays based on a set of events (e.g., acoustic events) with unknown event time. A localization algorithm may iteratively run to compute the acoustic event times, the observation delays, and the relative locations of the events and the sensors.
1. A method comprising:
measuring observation times of a set of events at a set of sensors;
generating initial estimates for internal delays of the sensors and event times of the events;
performing an iterative approximation algorithm to find a rank-3 approximation of the internal delays and event times using a criterion; and
computing locations of the sensors using the rank-3 approximation of the internal delays and event times.
2. The method of claim 1 , wherein the set of events is a set of acoustic events, and the set of sensors is a set of microphones.
3. The method of claim 2 , wherein one or more of the acoustic events is generated by human with hand claps or speech.
4. The method of claim 2 , wherein one or more of the acoustic events is generated by a device with a loudspeaker.
5. The method of claim 4 , wherein the device is a telephone.
6. The method of claim 1 , wherein one or more of the sensors are microphones located on mobile telephones.
7. The method of claim 1 , wherein the localization of the sensors is for the purpose of near-field beam-forming.
8. The method of claim 1 , wherein the localization of the sensors is used to identify talkers in a conference call.
9. The method of claim 1 , wherein the localization of the sensors is used as a basis for the enhancement of recorded audio signals.
10. The method of claim 1 , further comprising using the locations of the sensors to determine locations of one or more mobile telephones.
11. The method of claim 1 , further comprising computing locations of the acoustic events using the rank-3 approximation of the internal delays and event times.
12. The method of claim 1 , wherein the iterative approximation algorithm converges to a global solution for the internal delays and the event times.
13. The method of claim 12 , further comprising using the global solution as an initial value for an iterative maximum-likelihood algorithm to find a global minimum.
14. The method of claim 12 , further comprising using the global solution as an initial value for an iterative maximum-likelihood algorithm to minimize a squared error criterion.
15. The method of claim 1 , wherein the criterion is selected to facilitate finding a global solution.
16. The method of claim 1 , wherein the criterion is the Frobenius norm.
17. The method of claim 1 , wherein each step of the iterative approximation algorithm minimizes the criterion.
18. The method of claim 17 , wherein the criterion is the Frobenius norm.