Sound Source Separation Matrix Initialization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound source separation systems, such as those described in Japanese Patent No. 4444345, fail to accurately separate sound sources when the sound source changes, as the separation matrix does not converge to an optimal value, leading to ineffective signal separation.
Innovation Solution
A sound source separation apparatus and method that includes a transfer function storage unit, a sound change detection unit, and a parameter selection unit to calculate an initial separation matrix based on change state information, allowing for effective separation of sound sources even when the sound source changes, using a cost function that combines separation sharpness and geometric constraint functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the separation matrix is continuously updated using gradient descent, then the separation precision is improved under stable conditions, but the system cannot adapt when the sound source changes
Solution Approach 1:
The system dynamically switches between two operational modes: continuous gradient descent updates for stable conditions, and initialization from transfer function storage when sound source changes are detected. This dynamic adaptation allows the separation matrix to maintain precision while becoming responsive to environmental changes.
Solution Approach 2:
A sound source change detection mechanism provides feedback to the separation matrix update process. When changes are detected, the system receives feedback to reinitialize the separation matrix from stored transfer functions, creating a closed-loop system that adapts to changing acoustic environments.
2Adaptability or versatility
If the separation matrix is reinitialized frequently to adapt to sound source changes, then adaptability is improved, but separation precision deteriorates due to insufficient convergence
Solution Approach 1:
Transfer functions are pre-calculated and stored in advance for various sound source positions. When a sound source change is detected, the system immediately retrieves the pre-computed separation matrix from storage rather than initializing from scratch, allowing rapid adaptation without sacrificing precision through insufficient convergence.
3Speed
If the separation matrix is updated with large steps to respond quickly to changes, then response speed is improved, but separation precision deteriorates
Solution Approach 1:
Separation matrices for various sound source positions are pre-computed and stored in advance. When changes are detected, the system retrieves the appropriate pre-computed matrix immediately, achieving fast response without the precision loss that would result from large-step iterative updates.
Solution Approach 2:
The system creates copies of separation matrices for different sound source positions and stores them in advance. Upon detecting a sound source change, it copies the appropriate pre-computed matrix rather than performing iterative updates, achieving both speed and precision.
Data Source
AI summary
A sound source separation apparatus includes a transfer function storage unit that stores a transfer function from a sound source, a sound change detection unit that generates change state information indicating a change of the sound source on the basis of an input signal input from a sound input unit, a parameter selection unit that calculates an initial separation matrix on the basis of the change state information generated by the sound change detection unit, and a sound source separation unit that separates the sound source from the input signal input from the sound input unit using the initial separation matrix calculated by the parameter selection unit.