Noise Reduction Weighting for Headset Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise reduction methods for headsets struggle to maintain consistent performance during both speech and speech-absent periods, with SNR-based methods providing insufficient noise reduction during gaps.
Innovation Solution
The method involves receiving audio signals from internal and external microphones, processing them to obtain full-band power estimates, and adjusting weights based on these estimates to blend the signals, prioritizing the signal with lower full-band power during speech gaps to maintain 20-30 dB noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If SNR-based methods are used for noise reduction, then noise reduction is effective during speech periods, but noise reduction performance becomes insufficient during speech gaps
Solution Approach 1:
The system dynamically switches between full-band power estimation and SNR-based weighting based on speech detection. During speech gaps, full-band power estimation is used to maintain noise reduction consistency, while during speech periods, SNR-based weighting optimizes noise reduction effectiveness. This dynamic adaptation resolves the contradiction between consistent noise reduction and effective noise reduction during different periods.
Solution Approach 2:
The system changes the weighting parameter selection based on speech presence detection. When speech is absent, full-band power estimates are used to determine weights, ensuring consistent noise reduction. When speech is present, SNR-based weights are applied for optimized noise reduction. This parameter change strategy allows the system to maintain both consistency and effectiveness across different operational states.
2Object-affected harmful factors
If internal microphone is used for acoustic isolation, then noise reduction is achieved, but voice quality becomes muffled and bandwidth is reduced
Solution Approach 1:
The system merges signals from both internal and external microphones through adaptive weighting. The internal microphone provides noise isolation while the external microphone provides natural voice quality. By combining both signals with appropriate weights based on full-band power estimates, the system achieves both noise reduction and voice quality preservation simultaneously.
Solution Approach 2:
The system applies different quality characteristics to different frequency bands and time periods. During speech gaps, full-band power estimation emphasizes noise reduction. During speech periods, the system balances noise reduction with voice quality preservation. This local quality adjustment allows the system to optimize for different requirements in different contexts.
3Reliability
If full-band power estimation is used for weight adjustment, then noise reduction consistency is improved, but computational complexity increases
Solution Approach 1:
The system performs full-band power estimation periodically rather than continuously. It detects speech gaps using periodic power analysis and only applies full-band power-based weighting during these gaps. During normal speech periods, the system uses lighter SNR-based weighting. This periodic application reduces computational complexity while maintaining noise reduction consistency when needed.
Data Source
AI summary
Methods and systems for providing consistency in noise reduction during speech and non-speech periods are provided. First and second signals are received. The first signal includes at least a voice component. The second signal includes at least the voice component modified by human tissue of a user. First and second weights may be assigned per subband to the first and second signals, respectively. The first and second signals are processed to obtain respective first and second full-band power estimates. During periods when the user's speech is not present, the first weight and the second weight are adjusted based at least partially on the first full-band power estimate and the second full-band power estimate. The first and second signals are blended based on the adjusted weights to generate an enhanced voice signal. The second signal may be aligned with the first signal prior to the blending.