We incorporate deep learning techniques into capacitive images of body parts (ear, four fingers, and thumb) to improve the performance of user authentication in smartphones. Use of a capacitive touchscreen as an image sensor has several advantages, such as it is less sensitive to poor illumination conditions, occlusions, and pose variations. Also, it does not need an additional hardware like iris or fingerprint scanner. Use of capacitive images for user authentication is not new. However, the performance, specially, false reject rates (FRRs) of the state-of-the-art capacitive image-based systems are poor. In this paper, we focus on improving the performance and leverage deep learning. Deep learning techniques demonstrated spectacular performance in previous physical biometrics-based research. However, to our knowledge, effectiveness of deep learning is still unexplored in capacitive touchscreen-based user authentication. In order to bridge this research gap, we devise a multi-modal deep learning model, namely UASNet, and compare its performance with a large set of uni- and multi-modal baselines. Using the UASNet, we achieve an accuracy of 99.77%, an EER of 0.48%, and an FRR of 1.19% at FAR of 0.06%. 

Incorporating deep learning into capacitive images for smartphone user authentication

Continuous authentication (CA) is a promising approach to authenticate workers and avoid security breaches in the industry, especially in Industry 4.0, where most interaction between workers and devices takes place. However, introducing CA in industries raises the following unsolved questions regarding machine learning (ML) models: its precision and performance; its robustness; and the issue about if or when to retrain the models. To answer these questions, this article explores these issues with a proposed supervised versus nonsupervised ML-based CA system that uses sensors, applications statistics, or speaker data collected by the operator’s devices. Experiments show supervised models with equal error rates of 7.28% using sensors data, 9.29% with statistics, and 0.31% with voice, a significant improvement of 71.97, 62.14, and 97.08%, respectively, over unsupervised models. Voice is the most robust dimension when adding new workers, with less than 2% of false acceptance rate even if workforce size is doubled.

A Supervised ML Biometric Continuous Authentication System for Industry 4.0

Continuous authentication (CA) is a promising approach to authenticate workers and avoid security breaches in the industry, especially in Industry 4.0, where most interaction between workers and devices takes place. However, introducing CA in industries raises the following unsolved questions regarding machine learning (ML) models: its precision and performance; its robustness; and the issue about if or when to retrain the models. To answer these questions, this article explores these issues with a proposed supervised versus nonsupervised ML-based CA system that uses sensors, applications statistics, or speaker data collected by the operator’s devices. Experiments show supervised models with equal error rates of 7.28% using sensors data, 9.29% with statistics, and 0.31% with voice, a significant improvement of 71.97, 62.14, and 97.08%, respectively, over unsupervised models. Voice is the most robust dimension when adding new workers, with less than 2% of false acceptance rate even if workforce size is doubled. 

/pdf/a-supervised-ml-biometric-continuous-authentication-system-5d1eejya.pdf

Investigating the role of gesture modalities and screen size in an AR 3D game

As more people rely on smartphones to store sensitive information, the need for robust security measures is all the more pressing. Because traditional one shot authentication methods like PINs and passwords are vulnerable to various attacks, we present a behavioral biometrics based smartphone authentication system using swipes. While previous research focused on a single kind of swipe, our data set features swipes using different fingers and directions collected from 36 users across three sessions. In our system, we experimented with support vector machine (SVM) and random forest (RF) classifiers. We investigated which finger, direction, and classifier provided the best individual swipe authentication results. Then, we analyzed whether fusion of different fingers and directions improved results. The best unimodal result came from a rightward swipe with right thumb using SVM, which resulted in an area under ROC curve (AUC) of 0.936 and an equal error rate (EER) of 0.135. We found that swipes using thumbs offered better performance. Fusion improves results for the most part, and our best result was the combination of a leftward swipe with right thumb and a leftward swipe with left thumb. This combination gave an AUC of 0.969 and EER of 0.081 with the SVM classifier. 

Leran Wang

Papers

The effectiveness of zoom touchscreen gestures for authentication and identification and its changes over time