Pop Noise Control Mask for Acoustic Echo Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing acoustic echo cancellation systems and noise reduction approaches fail to effectively control impulsive noise, known as pop-noise or transient noise, which can drive loudspeakers beyond their linear range, leading to nonlinear components that cannot be removed and result in strong remaining impulsive parts in error signals.
Innovation Solution
A pop noise control system that calculates and applies a pop noise removal mask by analyzing spectral signal-to-noise ratios in a specific frequency range to distinguish and suppress impulsive noise parts, while minimizing the impact on speech signals, using a combination of spectral and temporal smoothing stages and background noise estimation to create a mask that is applied to the error signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If common acoustic echo cancellation or noise reduction approaches are applied, then general noise is reduced, but impulsive pop-noise components remain strong and unremoved
Solution Approach 1:
The patent applies different processing strategies to different parts of the signal spectrum. A pop-noise removal mask is created specifically for low-frequency bands (below a predetermined upper frequency limit) where pop-noise typically occurs, while other frequency regions are processed using conventional noise reduction methods. This localized approach allows effective pop-noise suppression without adversely affecting other frequency components.
Solution Approach 2:
The patent changes the processing parameters dynamically based on signal characteristics. By detecting impulsive components through signal-to-noise ratio analysis and determining their temporal and spectral properties, the system adjusts the pop-noise removal mask parameters (frequency limit, threshold values, smoothing factors) to adapt to varying pop-noise conditions while preserving speech quality.
2Object-generated harmful factors
If aggressive noise filtering is applied to remove impulsive components, then pop-noise is reduced, but speech quality deteriorates
Solution Approach 1:
The patent employs dynamic adaptation in the pop-noise removal process. The upper frequency limit for pop-noise processing, the signal-to-noise ratio thresholds, and the smoothing parameters are adjusted dynamically based on the detected signal characteristics and population of impulsive components. This dynamic approach ensures effective pop-noise removal while adapting to preserve speech components that may occur in similar frequency ranges.
Solution Approach 2:
The system uses feedback from signal-to-noise ratio estimation and impulsive component detection to control the pop-noise removal mask application. By continuously monitoring the error signal and reference signal characteristics, the system adjusts the masking parameters in real-time, applying stronger suppression when pop-noise is detected and reducing suppression when speech components are present, thereby maintaining speech quality.
3Object-affected harmful factors
If spectral processing is applied across the full frequency range, then all noise components are addressed, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the frequency spectrum into different processing zones. The low-frequency region below the predetermined upper frequency limit is processed with the specialized pop-noise removal mask, while higher frequency regions use conventional noise reduction processing. This segmentation reduces computational complexity by applying complex processing only where necessary (low frequencies where pop-noise occurs) rather than across the entire spectrum.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
Example pop noise removal systems and methods include detecting impulsive components in an input signal based on a signal-to-noise ratio spectrum of the input signal, and generating a spectral pop noise removal mask and applying the spectral pop noise removal mask to the input signal if impulsive components in the input signal are detected, the pop noise removal mask configured to suppress the impulsive components in the input signal, when applied.