Dual Microphone Noise Suppression via Beamforming and Blind Source Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise reduction solutions in noisy environments are ineffective in suppressing all noise sources independent of their temporal characteristics, location, or movement, often producing significant audio artifacts or failing to isolate speech from background noise effectively.
Innovation Solution
A system and method utilizing omnidirectional microphones with phase delay filtering to form speech and noise beamforms, combined with blind source separation and dual input spectral subtraction noise suppression, effectively isolates and reduces noise from speech signals, enhancing speech quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional noise reduction solutions are used, then certain types of noise can be suppressed, but they produce significant audio artifacts or fail to isolate speech from background noise effectively
Solution Approach 1:
The system segments the acoustic signal into distinct components (speech and noise) using beamforming techniques that spatially separate sound sources. Two omnidirectional microphones capture the mixed signal, and phase delay filtering divides it into a speech beamform and a noise beamform based on directional characteristics, enabling selective processing of each component.
Solution Approach 2:
The system extracts the noise component from the mixed acoustic signal by using the noise beamform, which is specifically designed to isolate noise sources. The extracted noise estimate is then used by the spectral subtraction module to remove noise from the speech signal without affecting the speech content.
2Adaptability or versatility
If existing noise suppression methods are applied, then some noise types can be reduced, but they fail to suppress all noise sources independent of their temporal characteristics, location, or movement
Solution Approach 1:
The system achieves universal noise suppression by combining multiple processing techniques: beamforming for spatial separation, blind source separation for statistical independence, and dual-input spectral subtraction for spectral analysis. This multi-functional approach handles various noise types (stationary, non-stationary, moving, stationary) uniformly without requiring separate processing paths.
Solution Approach 2:
The system dynamically adapts to changing noise conditions by continuously updating the noise estimate through spectral subtraction and using adaptive filtering techniques. The beamforming coefficients and spectral subtraction parameters are adjusted in real-time based on the instantaneous signal characteristics, enabling consistent performance across different temporal scenarios.
3Measurement precision
If directional filtering and beamforming are used to isolate speech, then speech isolation improves, but the system complexity increases
Solution Approach 1:
The system uses two omnidirectional microphones instead of a complex array of directional microphones. This partial approach (using minimal microphones) combined with post-processing beamforming achieves speech isolation without the mechanical complexity of directional microphone arrays, balancing performance and simplicity.
Data Source
Figure 1~2
Figure 3~5
Figure 6~7
AI summary
A system, method, and apparatus for separating speech signal from a noisy acoustic environment. The separation process may include directional filtering, blind source separation, and dual input spectral subtraction noise suppressor. The input channels may include two omnidirectional microphones whose output is processed using phase delay filtering to form speech and noise beamforms. Further, the beamforms may be frequency corrected. The omnidirectional microphones generate one channel that is substantially only noise, and another channel that is a combination of noise and speech. A blind source separation algorithm augments the directional separation through statistical techniques. The noise signal and speech signal are then used to set process characteristics at a dual input noise spectral subtraction suppressor (DINS) to efficiently reduce or eliminate the noise component. In this way, the noise is effectively removed from the combination signal to generate a good qualify speech signal.