Speech Quality Evaluation System Noise Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech quality evaluation systems struggle to accurately predict subjective opinion scores, especially when noise is present, as they fail to account for the influence of noise on speech quality and only provide a limited scale for evaluation, which is not sufficient for assessing phone speech quality effectively.
Innovation Solution
A speech quality evaluation system that calculates speech distortion by subtracting frequency characteristics of noise from the evaluation speech, using a noise characteristics calculation unit to obtain frequency characteristics during silence and speech durations, and a subjective evaluation prediction unit to calculate predicted subjective opinion scores based on these distortions, allowing for multiple scales of evaluation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If psychological experiments with multiple evaluators are conducted to assess speech quality, then evaluation accuracy is improved, but time consumption and cost increase
Solution Approach 1:
The patent creates an objective evaluation system that copies the functionality of human evaluators by using algorithmic processing of speech signals. Instead of relying on multiple human subjects, the system uses automated calculations based on speech data to predict mean opinion scores, thereby achieving evaluation accuracy without the time and resource costs of psychological experiments
Solution Approach 2:
The patent replaces the mechanical system of human psychological evaluation with an automated computational system. By substituting human evaluators with algorithmic processing that analyzes speech signals and calculates quality metrics, the system eliminates the need for time-consuming manual assessment while maintaining evaluation capability
2Productivity
If existing objective evaluation methods compare reference speech and far-end speech to predict subjective opinion scores, then evaluation efficiency is improved, but prediction accuracy deteriorates when noise is present
Solution Approach 1:
The patent extracts and separately analyzes the noise component from the speech signal by identifying periods where only noise is present. By isolating the noise characteristics and removing its influence from the quality assessment, the system maintains prediction accuracy even when noise is present in the evaluation speech
Solution Approach 2:
The patent applies different evaluation criteria to different portions of the speech signal. By identifying segments containing only noise and segments containing speech, the system applies localized analysis methods that account for the presence or absence of speech content, thereby improving overall prediction accuracy in noisy environments
3Device complexity
If only a limited scale for speech quality evaluation is provided, then system complexity is reduced, but evaluation comprehensiveness deteriorates
Solution Approach 1:
The patent creates a universal evaluation framework that can handle multiple evaluation scales and conditions through a single system architecture. By designing the system to accommodate different speech types (with or without noise) and multiple quality assessment dimensions, it achieves comprehensive evaluation capability without proportionally increasing system complexity
Data Source
AI summary
In prediction of a speech quality evaluation score such as a phone speech, even when a background noise exists, a subjective opinion score is predicted with high precision. A speech quality evaluation system that outputs a predicted value of the subjective opinion score for an evaluation speech such as a far-end speech of a phone, includes a speech distortion calculation unit that conducts, after calculating frequency characteristics of the evaluation speech, a process of subtracting given frequency characteristics from frequency characteristics of the evaluation speech, and calculates the speech distortion on the basis of the frequency characteristics after the subtracting process has been conducted, and a subjective evaluation prediction unit that calculates the predicted value of the subjective opinion score on the basis of the speech distortion.