Neural Network Hearing Aid Audio Path Switching for Noisy Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hearing aids struggle to effectively separate speech from background noise in noisy environments due to limitations in computational power and the impracticality of incorporating neural networks, leading to decreased speech intelligibility for individuals with hearing loss.
Innovation Solution
A dual-path signal processing system in hearing aids that integrates a neural network engine (NNE) and a digital signal processor (DSP), allowing selective engagement of neural network-based audio enhancement, with a controller determining the processing path based on user inputs, environmental factors, and sensor data to optimize audio processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural network algorithms are integrated into hearing aids to separate speech from background noise, then speech intelligibility in noisy environments is improved, but power consumption increases and battery life is reduced
Solution Approach 1:
The patent divides the neural network processing into multiple stages: a lightweight initial classification stage that runs continuously with low power consumption, and a more intensive speech separation stage that is only activated when speech is detected. This segmentation allows the system to maintain speech intelligibility improvements while significantly reducing overall power consumption compared to running full neural network processing continuously.
Solution Approach 2:
The system employs periodic activation of the neural network based on environmental conditions and speech detection events rather than continuous operation. The controller monitors acoustic environment parameters and selectively engages the neural network processor only when beneficial, creating a periodic action pattern that reduces average power consumption while maintaining speech intelligibility when needed.
2Adaptability or versatility
If neural network processing is continuously activated to handle varying acoustic environments, then adaptability is improved, but latency increases and real-time performance deteriorates
Solution Approach 1:
The system performs preliminary acoustic environment analysis using lightweight processing to detect speech presence and characterize the acoustic scene before activating the full neural network speech separation algorithm. This preliminary action allows the system to prepare for speech processing in advance, reducing the actual speech separation latency while maintaining adaptability to varying acoustic environments.
Solution Approach 2:
The patent implements dynamic control of the neural network processing pipeline, where the controller adjusts the activation and processing parameters based on real-time speech detection and acoustic environment assessment. This dynamic approach allows the system to maintain high adaptability to different environments while minimizing latency by only processing at full intensity when speech is actually present and separation is beneficial.
3Measurement precision
If advanced neural network algorithms are implemented to handle complex noise separation, then speech intelligibility is improved, but device complexity and computational requirements increase
Solution Approach 1:
The neural network processing is segmented into a simple speech detection component that runs on the main processor with low computational requirements, and a more complex speech separation component that runs on a specialized neural network processor. This segmentation allows advanced algorithms to be used for separation while keeping the overall device complexity manageable by distributing computational loads appropriately.
Solution Approach 2:
The patent introduces a controller as an intermediary that manages the interaction between the main processor and the neural network processor. This intermediary coordinates processing tasks, manages data flow, and controls activation timing, thereby reducing the computational burden on individual processors and simplifying the overall system architecture while maintaining high speech intelligibility performance.
4Power
If directional microphones and beamforming techniques are used to increase signal-to-noise ratio, then speech separation is improved, but accuracy decreases due to incorrect assumptions about speaker position and signal characteristics
Solution Approach 1:
The patent replaces traditional mechanical signal processing approaches (directional microphones and beamforming that rely on spatial assumptions) with neural network-based acoustic scene analysis and speech separation. This substitution eliminates the need for incorrect assumptions about speaker position and signal characteristics, using instead data-driven models that adapt to actual acoustic environments, thereby improving speech separation accuracy while maintaining signal-to-noise ratio improvements.
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.