Adaptive Filter Array for Speech Signal Noise Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech communication and recognition systems face challenges in effectively enhancing target audio signals while suppressing unwanted interference and noise, particularly in environments with multiple sources and complex geometries, leading to degraded signal quality.
Innovation Solution
The system incorporates a third adaptive filter and employs novel signal processing methods, including adaptive spatial filtering, frequency domain processing, and noise suppression techniques to isolate and enhance target audio signals, using multiple microphones and digital signal processing to separate target signals from interference and noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If adaptive filtering and noise suppression techniques are applied to enhance target audio signals, then signal quality is improved, but device complexity increases
Solution Approach 1:
The signal processing system is divided into multiple independent adaptive filters (first adaptive filter for target signal enhancement, second adaptive filter for interference suppression, third adaptive filter for noise suppression) that process signals in sequence. Each filter operates as a separate module with specific functionality, allowing complex signal processing to be achieved through composition of simpler individual filter operations.
Solution Approach 2:
The patent transforms the signal processing from time-domain only to frequency-domain processing by applying Fast Fourier Transform (FFT) after the adaptive filtering stages. This dimensional transition to frequency domain enables additional noise suppression capabilities and spectral analysis that cannot be achieved in the time domain alone, thereby improving signal quality without proportionally increasing time-domain computational complexity.
2Measurement precision
If multiple adaptive filters are incorporated to suppress interference and noise, then speech recognition accuracy is improved, but processing time increases
Solution Approach 1:
The adaptive filters operate in discrete processing blocks or frames of audio data, where each block is processed independently through the sequence of adaptive filters and FFT operations. This periodic batch processing allows for optimized computational routines within each block and enables parallel processing of multiple blocks, reducing overall processing time while maintaining high speech recognition accuracy through cumulative filtering effects.
3Measurement precision
If frequency domain processing is applied to further suppress noise, then signal quality is improved, but computational complexity increases
Solution Approach 1:
The first and second adaptive filters perform preliminary signal enhancement and interference suppression in the time domain before the signal is transformed to the frequency domain. This preliminary processing reduces the amount of noise and interference that needs to be handled in the computationally intensive FFT stage, thereby improving overall signal quality while mitigating the computational complexity burden of frequency domain processing.
Data Source
AI summary
The present invention uses a method of processing signals in which signals received from an array of sensors are subject to system having a first adaptive filter arranged to enhance a target signal and a second adaptive filter arranged to suppress unwanted signals. The output of the second filter is converted into the frequency domain, and further digital processing is performed in that domain. The invention is further enhanced by incorporating a third adaptive filter in the system and a novel method for performing improved signal processing of audio signals that are suitable for speech communication.