Hearable Active Acoustic Sensing for Noise-Resistant Speech Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless hearables face challenges in performing accurate speech recognition in noisy environments and socially inappropriate quiet environments without requiring hardware changes, and existing voice control technologies struggle with noise interference and user inconvenience.
Innovation Solution
The use of active acoustic sensing through ultrasound signals transmitted and received within the ear canal to perform speech recognition, which includes audioplethysmography, enabling silent speech recognition and enhancing recognition in noisy environments by fusing ultrasound and microphone data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If passive microphone sensing is used for speech recognition, then the device can capture audible speech, but it suffers from noise interference and poor performance in noisy environments
Solution Approach 1:
The patent introduces ultrasound signals as an intermediary medium to indirectly measure speech-related physiological changes. Instead of directly capturing audible speech through the microphone (which is susceptible to noise), the system transmits ultrasound signals through the ear canal and measures modifications to these signals caused by speech-related ear canal vibrations and muscle movements. This intermediary measurement approach isolates the speech detection process from external noise interference.
Solution Approach 2:
The patent replaces the acoustic/mechanical speech capture system (microphone listening to audible sound waves) with an ultrasound-based measurement system. By substituting the direct acoustic path with ultrasound transmission through tissue and fluid media, the system achieves speech recognition that is immune to external acoustic noise while still capturing the mechanical vibrations and movements associated with speech production.
2Measurement precision
If additional hardware is added to improve speech recognition in quiet environments, then recognition accuracy improves, but device complexity and cost increase
Solution Approach 1:
The patent makes the existing hearable device multi-functional by enabling it to perform both audio playback and speech recognition using the same ultrasound transmission and reception capabilities. The existing speakers and microphones are repurposed to generate and detect ultrasound signals, eliminating the need for dedicated speech recognition hardware while expanding the device's functionality into silent speech detection and physiological monitoring.
Solution Approach 2:
The hearable device serves itself by using its own existing components (speakers and microphones) to perform speech recognition functions. The device generates ultrasound signals through its speakers and detects the modified signals through its microphones, eliminating the need for external or additional specialized hardware. This self-service approach leverages existing resources to achieve new capabilities.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables discreet and accurate speech recognition in various environments, improving user experience by reducing noise susceptibility and eliminating the need for additional hardware, particularly benefiting users with motor speech disorders.
Implementation Method 1
performing speech recognition using active acoustic sensing... transmitting and receives at least one ultrasound signal, which propagates within a user's ear canal... This ultrasound signal can be modulated by a user's speech as well as by other muscle movements associated with speech
Data Source
AI summary
Techniques and apparatuses are described that perform speech recognition using active acoustic sensing. During active acoustic sensing, a hearable transmits and receives at least one ultrasound signal, which propagates within a user's ear canal. This ultrasound signal can be modulated by a user's speech as well as by other muscle movements associated with speech (e.g., jaw movement and/or tongue movement). As such, the ultrasound signal contains information that is correlated with speech as well as additional contextual information in how the user created the speech using their body. With active acoustic sensing, the hearable can directly perform speech recognition based on the ultrasound signal and/or enhance speech recognition by fusing information derived from the ultrasound signal with information derived from an audible signal that is passively sensed using a microphone.