Hearing Device EEG-Based Audio Processing for Cocktail-Party Effect
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hearing devices struggle to effectively help users focus on a specific speaker in a noisy environment with multiple speakers, known as the 'cocktail-party' effect, as existing technologies lack detailed methods to distinguish between speakers and adapt audio processing accordingly.
Innovation Solution
A hearing device equipped with brain-wave measurement sensors and a signal processing circuit that reconstructs audio spectra from EEG signals to determine the user's attention, adjusting audio processing to emphasize the attended speaker by altering the transfer function and acoustic directivity patterns, thereby enhancing speech intelligibility in noisy environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If brain-wave measurement sensors and EEG signal processing are added to hearing devices, then the ability to focus on specific speakers in noisy environments is improved, but device complexity increases
Solution Approach 1:
The patent introduces EEG sensors and brain-wave processing circuits as intermediary components between the user's cognitive state and the audio processing system. These intermediaries translate neural attention signals into actionable audio enhancement parameters, enabling the device to automatically adapt to user focus without requiring manual control or complex user interaction.
Solution Approach 2:
The patent replaces traditional mechanical or manual audio adjustment mechanisms with neurophysiological signal-based control. Instead of using physical dials, buttons, or complex algorithms to detect user intent, the system directly measures brain wave patterns to determine attention focus, substituting mechanical interaction with biological signal processing.
2Reliability
If audio processing is adapted based on reconstructed audio spectra from EEG signals, then speech intelligibility is improved, but processing time and computational requirements increase
Solution Approach 1:
The patent performs preliminary reconstruction of audio spectra from EEG signals in advance of real-time audio processing. By pre-processing the neural signals to extract attention-related spectral information, the system prepares enhancement parameters ahead of time, reducing the computational burden during critical real-time audio delivery and minimizing processing delays.
Solution Approach 2:
The system implements a feedback loop where reconstructed audio spectra from EEG signals continuously inform and adjust the audio processing parameters. This closed-loop approach allows the device to adaptively refine speech enhancement based on real-time attention detection, improving intelligibility while optimizing processing efficiency through iterative adjustment rather than exhaustive computation.
Data Source
Figure 1
AI summary
Hearing device (1) arranged on or at least partly implanted in an individual's head and comprising: an input means (3,4,5,6) providing one or more input audio signals; a signal processing circuit (7,8) to process at least one of said one or more input audio signals to provide a processed audio signal; an output means (10) to provide an audible signal to the individual in dependence on the processed audio signal; two or more electrodes (13,14,20) to detect electric brain potentials of the individual; and a brain-wave measurement circuit (15,21,25) to determine one or more EEG signals from electric signals received from the two or more electrodes (13,14,20), wherein the hearing device (1) further comprises: a first spectrum analyser (11,12) to repeatedly determine first audio spectra of at least one signal from the group of signals consisting of said one or more audio input signals and the audible signal; a reconstructor (16) to repeatedly reconstruct second audio spectra from the one or more EEG signals; a first correlator (17,18) to repeatedly determine a first coherence between the first and the second audio spectra; and a control unit (19) to alter said processing in dependence on the first coherence.