Abnormal Frame Detection in Speech Signals via Energy and Singularity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting speech distortion in wireless communications are either costly and time-consuming, such as manual subjective tests, or lack timely automatic detection capabilities, which is essential for assessing speech quality during transmission.
Innovation Solution
An abnormal frame detection method that processes speech signals by dividing them into subframes, calculating local energy values, performing singularity analysis through wavelet decomposition, and determining frames as abnormal based on threshold criteria to identify distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual subjective test method is used to assess speech quality, then measurement precision is improved, but loss of time and productivity deteriorate
Solution Approach 1:
The patent replaces manual subjective testing with an automated computer-based detection system. The system uses digital signal processing techniques including energy calculation, zero-crossing rate analysis, and correlation analysis to automatically identify abnormal frames in speech signals, eliminating the need for human testers while maintaining assessment accuracy.
Solution Approach 2:
The system performs self-testing by automatically analyzing speech signals without requiring external human evaluation. The automated detection algorithm independently assesses speech quality by processing audio data through multiple computational steps, including frame-based energy calculation and pattern recognition, enabling the system to evaluate itself without human intervention.
2Productivity
If automated detection method is implemented, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The patent divides the speech signal into discrete frames and further segments each frame into multiple subframes. This segmentation allows the system to analyze local characteristics of different portions of the signal independently, improving both the speed of processing and the precision of distortion detection by identifying specific abnormal segments within the overall speech signal.
Solution Approach 2:
The system calculates local energy values and zero-crossing rates for each subframe individually, then compares these local characteristics against threshold values. This local quality analysis enables the system to detect subtle distortions in specific segments of the speech signal while maintaining overall processing efficiency through parallel evaluation of multiple segments.
3Measurement precision
If complex analysis methods are used to improve detection accuracy, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent performs preliminary calculations of energy values, zero-crossing rates, and correlation coefficients for each subframe before making the final abnormal frame determination. These preliminary analyses prepare the data in advance, allowing the system to use simple threshold comparison rules rather than complex real-time decision algorithms, thereby reducing overall system complexity while maintaining high detection accuracy.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
An abnormal frame detection method and apparatus are provided. The method includes: obtaining a signal frame from a speech signal, and dividing the signal frame into at least two subframes (301); obtaining a local energy value of a subframe of the signal frame, and obtaining, according to the local energy value of the subframe, a first characteristic value used to indicate a local energy trend of the signal frame (302); performing singularity analysis on the signal frame to obtain a second characteristic value (303); and determining the signal frame as an abnormal frame if the first characteristic value meets a first threshold and the second characteristic value meets a second threshold (304). It is implemented whether distortion occurs in a speech signal is detected.