Sound Source Localization Using Phase Spectrum Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the location of a sound source using microphones are not robust against interference such as noise and reverberation, particularly when relying on phase difference measurements or energy peak searches.
Innovation Solution
An array of microphones processes audio signals by dividing them into frames, performing phase analysis on each pair to generate candidate angles, tracking these angles over time, and selecting a final angle based on criteria such as phase distortion and presence score, which helps a mobile robot to reposition itself towards the sound source.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase difference measurement using cross correlation function is used, then sound source location can be determined, but the method is not robust against interference such as noise and reverberation
Solution Approach 1:
The patent transforms the acoustic signal from the time domain to the frequency domain by computing the frequency spectrum using Fast Fourier Transform (FFT). This parameter change allows the system to analyze phase differences at multiple frequency components simultaneously, making the localization more robust against noise and reverberation while maintaining accuracy.
Solution Approach 2:
The patent divides the audio signal into multiple frequency components through spectral analysis. By segmenting the signal into frequency bins and analyzing phase differences at each frequency, the system can distinguish between direct sound and reverberant components, improving robustness against interference while maintaining location accuracy.
2Measurement precision
If energy peak search method is used to determine sound source location, then location can be found, but the method is not robust against interference such as noise and reverberation
Solution Approach 1:
The patent introduces frequency spectrum analysis as an intermediary between the raw acoustic signal and the location determination. By computing the frequency spectrum and analyzing phase differences in the frequency domain, the system creates a more reliable representation of the sound source that is less susceptible to noise and reverberation interference.
Solution Approach 2:
The patent replaces the time-domain energy peak search method with a frequency-domain phase analysis method. This substitution uses spectral information and phase differences across frequency components to determine location, which is more robust against interference compared to direct time-domain energy maximization.
3Measurement precision
If multiple candidate angles are generated from phase analysis, then location accuracy can be improved, but the complexity of processing increases
Solution Approach 1:
The patent implements a feedback mechanism where candidate angles generated from phase analysis are fed back into the system for further processing. The system uses feedback from multiple candidate angles to refine the final location estimate, improving accuracy while managing complexity through iterative refinement rather than exhaustive search.
Solution Approach 2:
The patent generates multiple candidate angles from phase analysis at different frequency components, which is slightly excessive action. However, this partial over-generation allows the system to filter and select the most accurate candidate, improving final location accuracy while the complexity is managed by processing only the necessary candidates rather than all possible angles.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves the robustness against noise and reverberation, enabling accurate sound source localization and enhancing audio processing operations, such as face detection and interaction with human speakers.
Implementation Method 1
microphones that capture acoustic energy generated by that sound source
Implementation Method 2
compute frequency spectrum data for the signals using a Fast Fourier Transform (FFT)
Implementation Method 3
compute a cross power spectrum of the frequency spectrum data for the pair of signals
Implementation Method 4
extract a phase spectrum of the cross power spectrum. The phase spectrum indicates a phase difference between the signals
Data Source
AI summary
An array of microphones placed on a mobile robot provides multiple channels of audio signals. A received set of audio signals is called an audio segment, which is divided into multiple frames. A phase analysis is performed on a frame of the signals from each pair of microphones. If both microphones are in an active state during the frame, a candidate angle is generated for each such pair of microphones. The result is a list of candidate angles for the frame. This list is processed to select a final candidate angle for the frame. The list of candidate angles is tracked over time to assist in the process of selecting the final candidate angle for an audio segment.