Sound Source Localization Using Amplitude Difference Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound source localization methods face challenges in accurately determining the location of sound sources in reverberant environments using arrays of microphones, as they struggle to differentiate between amplitude differences caused by varying distances and microphone sensitivities.
Innovation Solution
The method calculates a frame amplitude difference vector based on short-time frame data from an array of microphones, evaluates its similarity to reference vectors from candidate locations, and estimates the sound source location using these similarities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing sound source localization methods are used in reverberant environments, then the system can operate in complex acoustic conditions, but the accuracy of determining sound source location deteriorates due to inability to differentiate amplitude differences caused by distance and microphone sensitivities
Solution Approach 1:
The patent segments the sound field analysis into multiple independent components: amplitude difference vectors, phase difference vectors, and their cross-product vectors. By dividing the complex localization problem into these separable vector components, the system can process each component independently and combine results, thereby maintaining accuracy in reverberant environments where traditional single-method approaches fail.
Solution Approach 2:
The patent transforms the localization approach by changing from single-parameter analysis to multi-parameter vector analysis. It introduces amplitude difference vectors, phase difference vectors, and cross-product vectors as new parameters, allowing the system to capture multiple aspects of sound field characteristics simultaneously. This parameter transformation enables differentiation between distance-related amplitude changes and sensitivity-related amplitude changes, resolving the accuracy deterioration in reverberant conditions.
2Measurement precision
If traditional amplitude-based localization methods are used, then the system can determine sound source position, but the precision deteriorates because amplitude differences cannot distinguish between distance effects and microphone sensitivity variations
Solution Approach 1:
The patent transitions from scalar amplitude analysis to vector-based analysis by introducing amplitude difference vectors that operate in an additional dimensional space. This vector representation adds directional and relational information that scalar amplitudes cannot provide, enabling the system to distinguish between distance-related variations and sensitivity-related variations through vector orientation and magnitude relationships.
Solution Approach 2:
The patent introduces cross-product vectors as intermediary elements that relate amplitude difference vectors and phase difference vectors. These cross-product vectors serve as mediators that capture the interaction between amplitude and phase information, providing additional constraints and information that help resolve the ambiguity between distance effects and sensitivity effects in the localization process.
Data Source
AI summary
Sound source localization apparatuses and methods are described. A frame amplitude difference vector is calculated based on short time frame data acquired through an array of microphones. The frame amplitude difference vector reflects differences between amplitudes captured by microphones of the array during recording the short time frame data. Similarity between the frame amplitude difference vector and each of a plurality of reference frame amplitude difference vectors is evaluated. Each of the plurality of reference frame amplitude difference vectors reflects differences between amplitudes captured by microphones of the array during recording sound from one of a plurality of candidate locations. A desired location of sound source is estimated based at least on the candidate locations and associated similarity. The sound source localization can be performed based at least on amplitude difference.