Impulsive Interference Suppression in Speech Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional single-channel noise suppression algorithms are ineffective in reducing impulsive interferences in noisy speech signals, particularly in environments with wind noise, as they often misidentify speech signals and reduce recognition performance.
Innovation Solution
A method that identifies high-energy components and their temporal derivatives, morphologically filters these derivatives to detect impulsive interference onsets, and estimates interference energies without relying on pitch frequency, using a spectral weighting framework to suppress interferences while protecting speech signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If standard single channel noise reduction algorithms are applied, then stationary noise suppression is improved, but speech recognition performance deteriorates when impulsive interferences are present
Solution Approach 1:
The patent changes the detection parameters from stationary noise characteristics to transient impulse characteristics. It uses temporal derivatives and morphological filtering to detect rapid energy changes that characterize impulsive interferences, rather than relying on spectral characteristics of stationary noise. This parameter transformation enables the system to distinguish between stationary noise and impulsive interference, resolving the contradiction by adapting the detection mechanism to the specific characteristics of the harmful factor.
2Device complexity
If conventional noise suppression algorithms are used, then processing simplicity is maintained, but effectiveness against impulsive interferences deteriorates
Solution Approach 1:
The patent segments the noise suppression process into distinct stages: detection phase using temporal derivatives and morphological filtering, and suppression phase using spectral subtraction. This segmentation allows the system to apply specialized processing only where needed (during impulsive interference) while maintaining simplicity during normal operation. The morphological filtering step divides the complex signal processing into manageable operations that can be implemented efficiently.
3Object-affected harmful factors
If pitch-based methods are employed, then wind noise suppression is improved, but computational complexity and pitch estimation difficulty increase
Solution Approach 1:
The patent extracts the essential characteristic of impulsive interference (rapid energy change) and isolates it from the complex pitch estimation process. By using temporal derivatives to detect energy changes and morphological filtering to identify interference patterns, the system extracts interference detection capability without requiring pitch frequency analysis. This extraction approach eliminates the need for complex pitch tracking algorithms while maintaining effective wind noise suppression.
4Object-affected harmful factors
If aggressive noise suppression is applied, then noise reduction is improved, but speech distortion increases
Solution Approach 1:
The patent implements feedback through the morphological filtering process, which continuously monitors the temporal derivatives and adjusts suppression levels based on detected interference patterns. The system provides feedback about the presence and characteristics of impulsive interference, enabling adaptive suppression that intensifies only when interference is detected. This feedback mechanism prevents aggressive suppression during clean speech segments, thereby reducing distortion while maintaining noise reduction effectiveness during interference events.
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
Methods and apparatus for reducing impulsive interferences in a signal, without necessarily ascertaining a pitch frequency in the signal, detect onsets of the impulsive interferences by searching a spectrum of high-energy components for large temporal derivatives that are correlated along frequency and extend from a very low frequency up, possibly to about several kHz. The energies of the impulsive interferences are estimated, and these estimates are used to suppress the impulsive interferences. Optionally, techniques are employed to protect desired speech signals from being corrupted as a result of the suppression of the impulsive interferences.