Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SonicID: User Identification on Smart Glasses with Acoustic Sensing

Ke Li, Devansh Agarwal, Ruidong Zhang, Vipin Gunda, Tianjun Mo, Saif Mahmud, Boao Chen, François Guimbretière, Cheng Zhang

TL;DR

A low-power and minimally-obtrusive system called SonicID, designed to authenticate users on glasses, which extracts unique biometric information from users by scanning their faces with ultrasonic waves and utilizes this information to distinguish between different users, powered by a customized binary classifier with the ResNet-18 architecture.

Abstract

Smart glasses have become more prevalent as they provide an increasing number of applications for users. They store various types of private information or can access it via connections established with other devices. Therefore, there is a growing need for user identification on smart glasses. In this paper, we introduce a low-power and minimally-obtrusive system called SonicID, designed to authenticate users on glasses. SonicID extracts unique biometric information from users by scanning their faces with ultrasonic waves and utilizes this information to distinguish between different users, powered by a customized binary classifier with the ResNet-18 architecture. SonicID can authenticate users by scanning their face for 0.06 seconds. A user study involving 40 participants confirms that SonicID achieves a true positive rate of 97.4%, a false positive rate of 4.3%, and a balanced accuracy of 96.6% using just 1 minute of training data collected for each new user. This performance is relatively consistent across different remounting sessions and days. Given this promising performance, we further discuss the potential applications of SonicID and methods to improve its performance in the future.

SonicID: User Identification on Smart Glasses with Acoustic Sensing

TL;DR

A low-power and minimally-obtrusive system called SonicID, designed to authenticate users on glasses, which extracts unique biometric information from users by scanning their faces with ultrasonic waves and utilizes this information to distinguish between different users, powered by a customized binary classifier with the ResNet-18 architecture.

Abstract

Smart glasses have become more prevalent as they provide an increasing number of applications for users. They store various types of private information or can access it via connections established with other devices. Therefore, there is a growing need for user identification on smart glasses. In this paper, we introduce a low-power and minimally-obtrusive system called SonicID, designed to authenticate users on glasses. SonicID extracts unique biometric information from users by scanning their faces with ultrasonic waves and utilizes this information to distinguish between different users, powered by a customized binary classifier with the ResNet-18 architecture. SonicID can authenticate users by scanning their face for 0.06 seconds. A user study involving 40 participants confirms that SonicID achieves a true positive rate of 97.4%, a false positive rate of 4.3%, and a balanced accuracy of 96.6% using just 1 minute of training data collected for each new user. This performance is relatively consistent across different remounting sessions and days. Given this promising performance, we further discuss the potential applications of SonicID and methods to improve its performance in the future.
Paper Structure (46 sections, 4 equations, 9 figures, 6 tables)

This paper contains 46 sections, 4 equations, 9 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Authentication on Non-wearable Devices
  4. Authentication on Wearable Devices
  5. Smart Glasses
  6. Other Wearable Devices
  7. Comparison between SonicID and Other Authentication Methods on Wearables
  8. Facial Biometirc Information Extraction with Active Acoustic Sensing
  9. Face-based Biometric Authentication
  10. Signal Transmission and Reception
  11. Echo Profile Calculation
  12. Machine Learning Model for Authentication
  13. Evaluation Metrics
  14. Prototype Design and Implementation
  15. MCU and Sensors
  16. ...and 31 more sections

Figures (9)

  • Figure 1: FMCW Signal Transmission, Reception and Processing.
  • Figure 2: Machine Learning Pipeline for SonicID.
  • Figure 3: Hardware Components and Prototype of SonicID.
  • Figure 4: Evaluation Setup and Average Results of 40 Participants across Sessions and Days.
  • Figure 5: Average TPR, FPR and BAC across 40 Participants with Different Amounts of Training Data.
  • ...and 4 more figures