Audio Processing Device SNR Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audio processing devices, such as hearing aids, face challenges in accurately estimating the signal-to-noise ratio (SNR) in noisy environments, particularly in determining the a priori SNR without direct access to target signal power spectral density, leading to suboptimal noise reduction and speech intelligibility.
Innovation Solution
The implementation of a recursive algorithm that uses non-linear smoothing and adaptive low-pass filtering to estimate the a priori SNR from the a posteriori SNR, with parameters optimized using supervised learning techniques, such as neural networks, to improve noise reduction and reduce musical noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional noise reduction algorithms are used in audio processing devices, then device complexity is reduced, but measurement precision of signal-to-noise ratio deteriorates
Solution Approach 1:
The patent introduces an intermediary variable (auxiliary variable) that represents the relationship between signal power and noise power. This auxiliary variable serves as a mediator that transforms the complex a priori SNR estimation problem into a more manageable form that can be solved using the available a posteriori SNR measurements and power spectral density estimates, thereby improving measurement precision without proportionally increasing device complexity
Solution Approach 2:
The patent replaces conventional mechanical/mathematical filtering approaches with a machine learning-based neural network model. The neural network is trained offline to learn the complex non-linear relationships in noisy speech signals, and during runtime, it provides accurate SNR estimates through forward propagation, substituting complex iterative mathematical optimizations with efficient neural network inference that achieves higher precision with manageable computational complexity
2Measurement precision
If a priori SNR estimation is performed without direct access to target signal power spectral density, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The patent employs feedback mechanisms where the estimated a priori SNR and auxiliary variables from previous time frames are fed back into the current estimation process. The recursive updates of auxiliary variables using past values and current measurements create a feedback loop that progressively refines the SNR estimation accuracy without requiring direct access to target signal power spectral density, thereby maintaining ease of operation while improving measurement precision
Solution Approach 2:
The patent performs preliminary offline training of neural network models using datasets with known ground truth SNR values. This preliminary action prepares the model with pre-learned knowledge about signal-noise relationships, enabling it to make accurate a priori SNR estimates during runtime without needing direct access to target signal power spectral density, thus achieving high precision while maintaining operational simplicity
Data Source
AI summary
An audio processing device comprises a) at least one input unit for providing a time-frequency representation Y(k,n) of an electric input signal representing sound consisting of target speech and noise signal components, where k and n are frequency band and time frame indices, respectively, b) a noise reduction system configured to b1) determine a first signal to noise ratio estimate γ(k,n) of said electric input signal, and to b2) determine a second signal to noise signal ratio estimate ζ(k,n) of said electric input signal from said first signal to noise ratio estimate γ(k,n) based on a recursive algorithm providing non-linear smoothing, and wherein a determination of said one or more bias and/or smoothing parameters comprises the use of supervised learning, e.g. one or more neural networks. The invention may be used in audio processing devices, such as hearing aids, headsets, ear phones, active ear protection systems, handsfree telephone systems, mobile telephones, etc.