Neural Network Hearing Aid Signal Processing for Noisy Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional hearing aids struggle to effectively enhance speech understanding in noisy environments due to limitations in computational power and the inability to robustly separate speech from background noise, especially in complex real-world scenarios, leading to decreased communication effectiveness for individuals with hearing loss.
Innovation Solution
Integration of a neural network engine in the signal processing chain of hearing devices, utilizing a dual-path signal processing system that selectively engages neural network-based enhancement and digital signal processing to dynamically adjust to environmental conditions, ensuring low latency and efficient power usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional digital signal processing techniques are used in hearing aids, then the device complexity and power consumption remain manageable, but the speech intelligibility in noisy environments deteriorates due to inability to robustly separate speech from background noise
Solution Approach 1:
The patent replaces traditional mechanical/digital signal processing methods with a neural network-based system. The neural network engine processes audio signals to separate speech from background noise, achieving superior speech intelligibility in noisy environments compared to conventional DSP techniques while managing device complexity through specialized hardware implementation.
2Reliability
If neural network algorithms are implemented in hearing aids, then speech separation from background noise improves, but power consumption increases beyond the limited battery capacity
Solution Approach 1:
The patent segments the signal processing task by implementing a dual-path architecture: a neural network engine handles complex speech separation tasks when needed, while a traditional digital signal processor handles routine processing. This segmentation allows the system to achieve superior speech separation capability while managing power consumption by activating the power-intensive neural network only when necessary.
Solution Approach 2:
The controller selectively activates the neural network engine based on environmental conditions and processing needs, rather than running it continuously. This periodic activation pattern allows the system to maintain speech separation capability while significantly reducing average power consumption compared to continuous neural network operation.
3Reliability
If neural network processing is applied to audio signals, then speech intelligibility enhances, but processing latency increases which degrades real-time conversation quality
Solution Approach 1:
The system performs preliminary processing through the digital signal processor for routine audio signals, and only activates the neural network engine when speech separation is actually needed based on environmental assessment. This preliminary filtering approach reduces average latency while maintaining speech intelligibility when required.
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.