Speech Enhancement via Multi-Device Microphone Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vehicle speech recognition systems face challenges due to high noise levels and limited microphone placement, which affects signal-to-noise ratio and speech separation, especially in multi-user environments.
Innovation Solution
A system comprising a time synchronization unit, a source separation unit, and a noise reduction unit using a neural network to synchronize and separate speech signals from multiple microphones, improving speech recognition by extracting features and reducing noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple microphones are used for speech separation, then speech separation performance improves, but device complexity and cost increase
Solution Approach 1:
The patent extracts and utilizes existing microphones from portable devices (mobile phones, tablets, wearables) that users already carry, rather than requiring a dedicated microphone array system. This approach achieves multi-source speech separation by leveraging microphones from multiple ad-hoc devices, thereby improving speech separation performance without increasing device complexity or requiring specialized hardware installations.
2Device complexity
If microphones are placed in fixed positions, then system simplicity is maintained, but speech recognition accuracy deteriorates in noisy environments
Solution Approach 1:
The patent transitions from fixed microphone placement to dynamic, flexible microphone positioning by utilizing portable devices that can be held or positioned by users in optimal locations. This dynamic approach allows microphones to be adaptively placed near different speakers' mouths, improving speech recognition accuracy in noisy environments while maintaining system simplicity through the use of existing portable devices.
3Device complexity
If conventional speech processing is used in multi-user environments, then system simplicity is maintained, but speech separation performance deteriorates
Solution Approach 1:
The patent applies universal speech separation algorithms that can handle both single-user and multi-user scenarios. The system processes speech signals from multiple portable devices using unified signal processing techniques that automatically adapt to the number of active speakers, achieving effective speech separation in multi-user environments without requiring complex user-specific configurations or separate processing paths.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and a system for speech enhancement are provided. The system includes a time synchronization unit configured to synchronize microphone signals sent from at least two microphones; a source separation unit configured to separate the synchronized microphone signals and output a separated speech signal, which corresponds to a speech source; and a noise reduction unit including a feature extraction unit configured to extract a speech feature of the separated speech signal and a neural network configured to receive the speech feature and output a clean speech feature. The present disclosure provides a solution which may realize the speech separation and noise reduction by using at least two microphones.