Neural Network-Enabled Hearing Aid Dual-Path Processing for Noisy Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional hearing aids struggle to effectively separate speech from background noise in noisy environments due to limitations in computational power and the inability to adapt to varying acoustic environments, leading to decreased speech intelligibility for individuals with hearing loss.
Innovation Solution
A neural network-enabled hearing aid with a dual-path signal processing chain that selectively engages a neural network and digital signal processor, allowing for real-time audio enhancement by isolating target sounds and adjusting to user preferences and environmental conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional digital signal processing techniques are used in hearing aids, then device complexity and power consumption are limited, but speech intelligibility in noisy environments deteriorates
Solution Approach 1:
The patent segments the signal processing task by dividing the audio spectrum into multiple frequency channels, with different processing paths for speech frequencies and non-speech frequencies. This allows complex neural network processing to be applied selectively only where needed (in speech frequencies), rather than to the entire audio spectrum, thereby reducing overall computational complexity while improving speech intelligibility in noisy environments.
2Reliability
If neural network algorithms are continuously engaged to separate speech from noise, then speech intelligibility improves, but battery life deteriorates
Solution Approach 1:
The patent implements periodic engagement of the neural network algorithm by continuously monitoring environmental noise levels and selectively activating the neural network processor only when noise exceeds a predetermined threshold. This periodic action allows the system to maintain speech intelligibility when needed while conserving battery life during quiet periods, directly addressing the contradiction between continuous processing and battery duration.
3Reliability
If neural network processing is applied to all audio frequencies, then noise reduction effectiveness improves, but power consumption increases
Solution Approach 1:
The patent applies local quality by differentiating processing strategies across different frequency bands. Neural network processing is applied specifically to speech-relevant frequency ranges where noise reduction is most critical for intelligibility, while other frequency ranges receive different or reduced processing. This localized approach maximizes noise reduction effectiveness in critical bands while minimizing overall power consumption.
Data Source
AI summary
The disclosure generally relates to a method, system and apparatus for processing audio through a neural network contained in a hearing device. 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; 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 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.