Adaptive Speech Enhancement for Intelligibility With Minimal Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech enhancement algorithms in noisy environments degrade audio quality with audible artifacts, compromising speech intelligibility, especially in critical communication scenarios like rescue operations.
Innovation Solution
A computer-implemented method optimizes speech enhancement parameters based on real-time intelligibility and quality targets, using a closed-form algorithm to adapt to noise conditions, minimizing artifacts and ensuring high speech intelligibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If speech enhancement processing is applied to improve speech intelligibility in noisy environments, then speech intelligibility is improved, but audible artifacts are introduced and audio quality is degraded
Solution Approach 1:
The patent optimizes parameters of a predetermined speech enhancement algorithm by calculating a measure of speech intelligibility and adjusting algorithm parameters to meet a speech intelligibility target while minimizing audio quality degradation. This involves changing parameters such as enhancement gain, filtering characteristics, and processing intensity to achieve the desired balance between intelligibility and quality.
Solution Approach 2:
The speech enhancement algorithm is made adaptive by continuously monitoring the acoustic environment and dynamically adjusting its parameters based on current noise conditions. The system calculates speech intelligibility measures in real-time and modifies enhancement strength accordingly, allowing it to provide strong enhancement when needed while reducing or eliminating enhancement when speech intelligibility is already sufficient, thereby minimizing artifacts.
2Measurement precision
If speech enhancement processing is applied to enhance speech intelligibility, then speech intelligibility is improved, but processing complexity and computational requirements increase
Solution Approach 1:
The patent employs a predetermined speech enhancement algorithm that has been pre-configured with optimized processing steps and parameter settings. By preparing the enhancement framework in advance with predetermined structures and optimization criteria, the system reduces real-time computational complexity while maintaining effective speech intelligibility enhancement capability.
Solution Approach 2:
The patent replaces complex iterative optimization mechanisms with a closed-form optimization algorithm that directly calculates optimal enhancement parameters based on measured speech intelligibility. This substitution of computational mechanics simplifies the processing requirements and reduces device complexity while achieving the same enhancement objectives.
3Measurement precision
If speech enhancement processing is applied to improve speech intelligibility in noisy environments, then speech intelligibility is improved, but power consumption increases
Solution Approach 1:
The patent applies speech enhancement processing selectively and adaptively based on measured speech intelligibility levels. Instead of continuously applying full-strength enhancement, the system adjusts the degree of processing to provide only the necessary enhancement to meet intelligibility targets. This partial action approach reduces unnecessary computational operations and lowers power consumption while maintaining adequate speech intelligibility.
4Measurement precision
If speech enhancement processing is applied to enhance speech intelligibility, then speech intelligibility is improved, but processing delay increases
Solution Approach 1:
The patent uses a predetermined speech enhancement algorithm with pre-configured processing steps and pre-optimized parameter sets. By having the enhancement framework prepared in advance with predetermined structures, the system minimizes real-time processing delays while maintaining effective speech intelligibility enhancement.
Solution Approach 2:
The patent implements a streamlined processing pipeline that rapidly calculates speech intelligibility measures and quickly adjusts enhancement parameters using closed-form solutions. This rushing through the optimization process with efficient algorithms reduces processing delay while achieving the necessary speech intelligibility enhancement.
Data Source
AI summary
A method for enhancement of speech intelligibility in a device arranged for a near-end side a communication with a far-end device. The method involves calculating a measure of speech intelligibility at the near-end side based on a near-end audio input and a far-end audio input. Then, based on the calculated measure of speech intelligibility optimizing parameters of a predetermined speech enhancement algorithm, where a predetermined speech intelligibility target, and an additional target are taken into account to generate an optimized speech enhancement algorithm. Next, processing the far-end audio input according to the optimized speech enhancement algorithm, and generating a near-end audio output accordingly. The algorithm can adapt to changing noise conditions and be optimized for both speech intelligibility and another target. This can be used to minimize delay, electric power consumption and audio quality while satisfying the speech intelligibility target. The optimization can be based on a closed-form solution.