Sound Direction Estimation via Source Class Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound direction estimation apparatuses, such as the one described in JP-A-2010-281816, cannot accurately estimate the direction of a sound source belonging to a specific class, as they perform processes regardless of sound classes.
Innovation Solution
A sound direction estimation apparatus that includes a sound source model storage unit to store likelihoods of sound source classes, a sound signal input unit, a sound source identification unit to identify the sound source class, and a sound source localization unit to estimate the sound direction of the identified class, using correlation matrices and eigenvectors calculated by the sound signal input and separation units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sound direction estimation is performed without classifying sound sources, then the processing is simpler and faster, but the estimation accuracy for specific sound classes deteriorates
Solution Approach 1:
The patent segments the sound direction estimation process into two distinct stages: first performing sound source classification to identify the type of sound, then performing direction estimation specific to that sound class. This segmentation allows the system to apply class-specific processing only when needed, improving accuracy for specific sound classes while maintaining reasonable processing complexity through the use of pre-built classification models.
2Measurement precision
If sound signals are processed as a mixture without separation, then the processing is simpler, but the direction estimation for individual sound sources deteriorates
Solution Approach 1:
The patent performs sound source separation as a preliminary action before direction estimation. By separating the mixed sound signals into individual source components first, the system can then estimate directions for each separated source independently. This preliminary separation improves individual source direction accuracy while managing processing complexity through efficient separation algorithms.
Data Source
AI summary
A sound direction estimation apparatus includes a sound source model storage unit that stores likelihood of a sound source class in correspondence with a sound feature quantity, a sound signal input unit that receives a sound signal, a sound source identification unit that identifies a sound source class of the sound signal input from the sound signal input unit with reference to the sound source model storage unit based on the sound feature quantity of the sound signal, and a first sound source localization unit that estimates a sound direction of the sound signal of the sound source class identified by the sound source identification unit.