Authentication Using Active Acoustic Sensing

Smart overview of the Invention

Techniques and apparatuses for authentication using active acoustic sensing are described, focusing on the use of hearables like wireless earbuds. These devices transmit and receive ultrasound signals within a person's ear canal. The ultrasound signals carry information related to vocalizations and contextual data on how these vocalizations are produced and transmitted through bone conduction. This method allows for the creation of an ultrasound-based voice signature, facilitating authentication with enhanced security metrics, such as target spoof acceptance and false acceptance rates.

Background

With the rise of wireless technology, hearables have become a common tool for accessing audio content while maintaining freedom of movement. As the market grows, there is a demand for adding features to hearables without altering their hardware. This patent addresses the need for enhanced security features, leveraging active acoustic sensing to authenticate users based on unique audio signatures derived from their vocalizations.

Authentication Process

Active acoustic sensing involves the transmission and reception of ultrasound signals that are influenced by a user's vocalizations and associated muscle movements. These signals, when processed, create a unique voice signature. This signature can be used alone or in combination with traditional voice signatures for authentication. The approach aims to maintain low spoof and false acceptance rates, thus improving security while allowing for seamless user interaction with electronic devices.

Advantages

This method offers several advantages over traditional authentication techniques. It enhances security by reducing the risk of spoofing, even in environments with background noise or when users speak quietly. The use of ultrasound signals helps distinguish between genuine and recorded voices, minimizing the chances of unauthorized access. Furthermore, it enables integration into existing hearables without the need for additional hardware, maintaining device aesthetics and functionality.

Implementation

The described system utilizes a hearable that forms an acoustic circuit within the ear, using ultrasound signals to detect physiological changes. This process, termed audioplethysmography, allows the hearable to perform authentication by recognizing changes in the acoustic circuit. The method supports various vocalizations, including speaking, whispering, and singing, providing a versatile and secure solution for user authentication in modern hearables.