Hearing Aid Speech Forecasting for Low-Latency Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio device systems, particularly hearing aids, face challenges in improving sound quality and reducing noise, with significant processing delays and power consumption limitations.
Innovation Solution
A method involving audio signal encoding, latent encoding manipulation, and probabilistic forecasting to predict future audio signals, distributed across computing devices to minimize delays and enhance sound quality, using autoencoders and forecasting models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If traditional hearing aid systems process audio signals locally, then device complexity is reduced, but noise reduction capability and sound quality improvement are insufficient
Solution Approach 1:
The audio processing system is divided into multiple segments: local processing in the hearing aid device and remote processing in an external device. The hearing aid performs initial audio capture and basic processing, while the external device handles complex noise reduction and speech enhancement algorithms, dividing the processing workload to achieve superior noise reduction without overburdening the hearing aid itself
Solution Approach 2:
A communication interface acts as an intermediary between the hearing aid and the external device. This intermediary enables the transfer of audio data and control signals, allowing the external device to receive audio information, process it through advanced algorithms, and send enhanced output back to the hearing aid, thereby achieving improved noise reduction while maintaining system manageability
2Object-affected harmful factors
If audio signal processing is performed with high computational power, then sound quality and noise reduction are improved, but processing delays increase
Solution Approach 1:
The processing pipeline is segmented into stages performed at different locations and times. The hearing aid performs immediate local processing for timely response, while the external device performs computationally intensive noise reduction in parallel. This segmentation allows high-quality processing without creating a single bottleneck that would cause excessive delays
Solution Approach 2:
The system performs preliminary audio capture and basic processing in the hearing aid before transferring to the external device. This preliminary action ensures that time-sensitive components are handled immediately, while the more time-flexible noise reduction processing occurs subsequently, maintaining overall processing speed while achieving high sound quality
3Measurement precision
If noise reduction algorithms are applied to audio signals, then speech intelligibility is improved, but processing time increases
Solution Approach 1:
The communication interface serves as an intermediary that enables efficient data exchange between devices. By optimizing this intermediary connection, the system can transfer audio data and processed results quickly, reducing the overall processing time while still allowing comprehensive noise reduction algorithms to run in the external device for improved speech intelligibility
Solution Approach 2:
The hearing aid applies partial noise reduction processing locally for immediate effect, while the external device applies more extensive algorithms. This partial action approach ensures that some speech intelligibility improvement is achieved quickly, while additional processing enhances the result further without causing excessive delays, as the hearing aid can operate independently at lower processing loads
Data Source
AI summary
A method of operating a hearing aid system in order to provide at least one of improved noise reduction and speech intelligibility and a hearing aid system adapted to carry out the method.