Audio Processing Device Adaptive SNR Smoothing
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 time-frequency domains, leading to inadequate noise reduction and potential musical noise issues.
Innovation Solution
The implementation of a recursive algorithm that uses non-linear smoothing, specifically low-pass filtering with an adaptive cut-off frequency, to estimate the a priori SNR from the a posteriori SNR, enhancing noise reduction by dynamically adjusting the filtering based on SNR conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-linear smoothing (low-pass filtering) is applied to a posteriori SNR estimate, then noise reduction performance is improved, but speech distortion and musical noise may increase
Solution Approach 1:
The patent implements dynamic adaptation of the smoothing parameter λ based on the estimated SNR value. When SNR is high, stronger smoothing is applied to reduce noise; when SNR is low, weaker smoothing is applied to preserve speech. This dynamic adjustment resolves the contradiction by making the smoothing intensity conditional on the acoustic situation, thereby reducing musical noise while maintaining noise reduction effectiveness.
Solution Approach 2:
The patent changes the smoothing parameter λ as a function of the SNR estimate. By adjusting this key parameter based on the acoustic environment, the system optimizes the balance between noise reduction and speech preservation, preventing musical noise artifacts that would result from fixed-parameter smoothing.
2Reliability
If strong smoothing is applied to SNR estimates, then noise reduction is improved, but speech clarity and transient response deteriorate
Solution Approach 1:
The smoothing strength is dynamically adjusted based on the SNR condition. In low-SNR regions, stronger smoothing is applied to improve noise reduction; in high-SNR regions, weaker smoothing preserves speech clarity and transient response. This dynamic approach resolves the contradiction by adapting smoothing intensity to the acoustic context.
Solution Approach 2:
Different smoothing strengths are applied locally based on the SNR characteristics of different time-frequency regions. This allows aggressive noise reduction in noise-dominated regions while preserving speech quality in speech-dominated regions, resolving the contradiction between noise reduction and speech clarity.
3Reliability
If adaptive low-pass filtering with variable cut-off frequency is used, then noise reduction performance is improved, but computational complexity increases
Solution Approach 1:
The cut-off frequency of the low-pass filter is adapted based on the SNR estimate. By dynamically changing this single parameter, the system achieves improved noise reduction performance without implementing a fully adaptive complex filter structure, thus managing computational complexity while enhancing performance.
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: determine a first signal to noise ratio estimate γ(k,n) of said electric input signal, and 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 parameters of said smoothing are determined in dependence of the first and/or the second signal to noise ratio estimates corresponding to a multitude of frequency band indices. The invention may be used in hearing aids, headsets, ear phones, active ear protection systems, handsfree telephone systems, mobile telephones, etc.