.png){kind=icon}
.png){kind=icon}
Return
User Authentication Method for Smart Rings Using Active Acoustic Sensing
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2025.3549278.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Wearable devices, such as bracelets and rings, have gained widespread popularity due to advancements in semiconductor technology, the miniaturization of devices and sensors, and the development of data analysis techniques. While tablets, smartwatches, and laptops incorporate biometrics like facial recognition and fingerprint identification for user authentication, many wearable devices lack user authentication features. In particular, smart rings (ring-type devices) do not come equipped with authentication capabilities. When user authentication is necessary, a fingerprint or an alternative device is required. This paper proposes an automatic and continuous user authentication method that utilizes active acoustic sensing, employing a speaker and microphone integrated within the smart ring. The proposed method authenticates the wearer by analyzing the frequency response of the acoustic signal captured by the microphone and comparing it to the frequency response of a pre-registered individual. The method takes advantage of the constant contact between the smart ring and the finger, as each person's finger has unique acoustic characteristics based on its shape and composition. Two types of devices were developed for this study: a tape device featuring a flexible piezoelectric element attached to Velcro tape, and a ring device with a flexible piezoelectric element mounted on a 3D-printed ring. The tape device was evaluated with seven participants in a relaxing hand state, and the observed average Equal Error Rate (EER) was 0.127. Additionally, the ring-shaped device was evaluated with thirteen participants in three different states: a relaxing hand, a gripping hand, and a typing hand. The observed average EER for the relaxing hand was 0.066, for the gripping hand was 0.026, and for the typing hand was 0.078.
Keywords:
Authentication
Acoustics
Sensors
Microphones
Accuracy
Hands
Piezoelectric transducers
Biometric authentication
Face recognition
Wearable devices
Active acoustic sensing
smart ring
user authentication
wearable computing