Dual-Path Neural Network Hearing Aid Processing for Speech Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional hearing aids struggle to effectively enhance speech in noisy environments due to limitations in computational power and the inability to adapt to varying acoustic environments, leading to challenges in isolating speech from background noise.
Innovation Solution
Integration of a neural network-based audio enhancement system within a hearing aid that selectively engages a dual-path signal processing chain, utilizing both digital signal processing and neural network processing to dynamically adjust to environmental noise and user preferences, ensuring optimal speech isolation and ambient noise management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural network processing is implemented in hearing aids, then speech isolation capability is improved, but power consumption increases
Solution Approach 1:
The system dynamically switches between dual-path processing (combining traditional DSP and neural network processing) and single-path processing (neural network only) based on environmental noise conditions. In noisy environments, both paths are used to maximize speech isolation. In quiet environments, only neural network processing is used to conserve power. This dynamic adaptation resolves the contradiction between speech isolation capability and power consumption.
Solution Approach 2:
The system changes the processing parameter (number of active processing paths) based on environmental conditions. When noise levels exceed a threshold, the system activates both traditional DSP and neural network processing paths. When noise levels are low, it switches to neural network only mode, thereby adjusting power consumption according to actual needs while maintaining speech isolation performance.
2Measurement precision
If dual-path signal processing is used, then speech enhancement is improved, but device complexity increases
Solution Approach 1:
The dual-path signal processing system dynamically adjusts which processing path is active based on environmental noise conditions. The controller monitors noise levels and switches between using both traditional DSP and neural network processing paths, or using only the neural network path. This dynamic approach improves speech enhancement when needed while reducing effective system complexity during normal operation.
3Measurement precision
If neural network processing is continuously active, then speech isolation is improved, but battery life decreases
Solution Approach 1:
Instead of continuously activating neural network processing, the system periodically assesses environmental noise conditions and activates neural network processing only when noise levels exceed a threshold. This periodic activation based on environmental conditions maintains speech isolation capability when needed while significantly extending battery life by avoiding unnecessary processing during quiet periods.
Solution Approach 2:
The system changes the operational state parameter of the neural network (active/inactive) based on environmental noise conditions. When noise levels are high, the neural network is activated to provide speech isolation. When noise levels are low, it is deactivated to conserve battery power. This parameter change strategy resolves the contradiction between speech isolation performance and battery life.
Data Source
AI summary
The disclosure generally relates to a method, system and apparatus to improve a user's understanding of speech in real-time conversations by processing the audio through a neural network contained in a hearing device. The hearing device may be a headphone or hearing aid. In one embodiment, the disclosure relates to an apparatus to enhance incoming audio signal. The apparatus includes a controller to receive an incoming signal and provide a controller output signal; a neural network engine (NNE) circuitry in communication with the controller, the NNE circuitry activatable by the controller, the NNE circuitry configured to generate an NNE output signal from the controller output signal; and a digital signal processing (DSP) circuitry to receive one or more of controller output signal or the NNE circuitry output signal to thereby generate a processed signal; wherein the controller determines a processing path of the controller output signal through one of the DSP or the NNE circuitries as a function of one or more of predefined parameters, incoming signal characteristics and NNE circuitry feedback.