Speech Enhancement Using Inter-Microphone Level Differences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise suppression methods in speech recordings, such as beamforming techniques, face challenges in accurately estimating time delays due to reverberation and require a large number of microphones, making them inefficient on small devices like cellular phones, especially at lower frequencies.
Innovation Solution
The method utilizes inter-microphone level differences (ILD) to differentiate between speech and noise by calculating energy estimates from multiple microphones, deriving noise estimates, and applying a Wiener filter to enhance speech, which is then converted back to the time domain for output.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If beamforming techniques are used for noise suppression, then spatial filtering capability is improved, but the number of microphones required increases significantly
Solution Approach 1:
The patent extracts and utilizes only the inter-microphone level difference (ILD) information from the microphone signals, rather than requiring full beamforming processing. By focusing specifically on the energy level differences between microphones, the system achieves noise suppression without needing a large microphone array
Solution Approach 2:
The patent changes the approach from using time delay estimates to using energy level differences (ILD) as the key parameter for noise suppression. This parameter change allows effective operation with fewer microphones, particularly at lower frequencies where time delay estimation becomes unreliable
2Volume of moving object
If beamforming is used on small devices, then portability is improved, but noise suppression effectiveness deteriorates at lower frequencies
Solution Approach 1:
The patent switches from time delay estimation to energy level difference (ILD) measurement as the operating parameter. ILD remains effective at low frequencies where time delay becomes ambiguous, allowing small devices to maintain noise suppression capability across the full frequency range
Solution Approach 2:
The patent replaces the mechanical/time-based approach of beamforming with an energy-based approach. By substituting time delay estimation with energy level comparison, the system achieves frequency-independent performance suitable for compact devices
3Reliability
If time delay estimation is used for source localization, then spatial filtering is improved, but measurement accuracy deteriorates due to reverberation
Solution Approach 1:
The patent extracts only the level difference information from the microphone signals, discarding the time delay estimation that is vulnerable to reverberation. By taking out only the robust ILD component, the system achieves spatial filtering without the measurement errors introduced by reverberant environments
Solution Approach 2:
The patent uses a simpler, more robust measurement approach (energy level comparison) that is less susceptible to environmental conditions. This disposable-like approach of using a simple, robust metric replaces the complex, fragile time delay estimation process
Data Source
AI summary
Systems and methods for utilizing inter-microphone level differences to attenuate noise and enhance speech are provided. In exemplary embodiments, energy estimates of acoustic signals received by a primary microphone and a secondary microphone are determined in order to determine an inter-microphone level difference (ILD). This ILD in combination with a noise estimate based only on a primary microphone acoustic signal allow a filter estimate to be derived. In some embodiments, the derived filter estimate may be smoothed. The filter estimate is then applied to the acoustic signal from the primary microphone to generate a speech estimate.