Acoustic Transfer Function Generation via Non-Discrete Direction Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating transfer functions in sound source localization and separation, such as those used in speech recognition, require extensive measurement and data collection for accurate angle resolution, limiting the ability to obtain transfer functions for arbitrary angles and requiring large databases, especially for high accuracy.
Innovation Solution
A transfer function generation apparatus and method that models acoustic transfer functions using a non-discrete argument for sound source arrival directions, allowing the generation of transfer functions for arbitrary angles through Fourier series expansion and the use of a Moore-Penrose pseudo-inverse matrix, enabling the creation of a database without prior measurement and reducing the need for equally spaced data points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If actual measurement-based transfer functions are collected for high accuracy sound source localization, then localization accuracy is improved, but the amount of time and effort required increases significantly
Solution Approach 1:
The patent creates a virtual copy of the physical measurement process by using a sound propagation model that replicates acoustic field characteristics. Instead of physically measuring transfer functions at every possible angle, the system generates a database by computationally modeling sound propagation from virtual sound sources at numerous positions, thereby obtaining accurate transfer functions without the time-consuming physical measurements
Solution Approach 2:
The patent changes the approach from physical measurement parameters to computational modeling parameters. By adjusting model parameters such as microphone positions, sound source positions, and acoustic field characteristics within the simulation, the system generates transfer functions that match real-world measurements while requiring minimal actual data collection time
2Adaptability or versatility
If transfer functions are measured at equally spaced angles to cover all directions, then completeness of the database is improved, but the number of required measurements increases
Solution Approach 1:
The patent segments the sound propagation problem by measuring transfer functions only at a limited set of discrete positions rather than continuously across all angles. The sound propagation model then interpolates between these discrete measurements to provide complete coverage, effectively dividing the task into manageable measurement points while maintaining comprehensive database coverage through computational assistance
Solution Approach 2:
The sound propagation model acts as an intermediary between the limited physical measurements and the complete database requirements. It takes the transfer functions measured at a few positions and computes the transfer functions at all other positions by modeling sound propagation physics, thereby bridging the gap between sparse measurements and comprehensive data needs
3Quantity of substance
If interpolation methods are used to obtain transfer functions at intermediate angles, then data requirements are reduced, but the ability to obtain arbitrary angle transfer functions is limited
Solution Approach 1:
The patent transforms the static interpolation approach into a dynamic computational model. Instead of using fixed interpolation formulas that work only for specific angle relationships, the sound propagation model dynamically calculates transfer functions at any arbitrary angle by simulating sound propagation physics, enabling flexible query of transfer functions at any position without being constrained by the measurement geometry
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 allows for accurate estimation of transfer functions at arbitrary angles with reduced data requirements, improving sound source localization and separation accuracy while minimizing the amount of data needed, and enabling the generation of transfer functions for non-equally spaced angles.
Implementation Method 1
a transfer function to each microphone of the microphone array is used in the sound source localization and the sound source separation. The transfer function is calculated by collecting a measurement signal that is output from the sound source using the microphone
Implementation Method 2
the transfer function is obtained by performing Fourier transform of the impulse response
Data Source
AI summary
A transfer function generation apparatus includes: a modeling part that models, using a function which uses an arrival direction of a sound source as a non-discrete argument, a plurality of acoustic transfer functions to a microphone from sound sources present in a plurality of directions and that stores the modeled function; and a transfer function generation part that generates a transfer function of an arbitrary direction by using the modeled and stored function.