Abstract
With the ubiquity of wearable devices, various behavioural biometrics have been exploited for continuous user authentication during daily activities. However, biometric authentication using complex hand behaviours have not been sufficiently investigated. This paper presents an implicit and continuous user authentication model based on hand-object manipulation behaviour, using a finger-and hand-mounted inertial measurement unit (IMU)-based system and state-of-the-art deep learning models. We employed three convolutional neural network (CNN)-based deep residual networks (ResNets) with multiple depths (i.e., 50, 101, and 152 layers) and two recurrent neural network (RNN)-based long short-term memory (LSTMs): simple and bidirectional. To increase ecological validity, data collection of hand-object manipulation behaviours was based on three different age groups and simple and complex daily object manipulation scenarios. As a result, both the ResNets and LSTMs models acceptably identified users’ hand behaviour patterns, with the best average accuracy of 96.31% and F1-score of 88.08%. Specifically, in the simple hand behaviour authentication scenarios, more layers in residual networks tended to show better performance without showing conventional degradation problems (the ResNet-152 > ResNet-101 > ResNet-50). In a complex hand behaviour scenario, the ResNet models outperformed user authentication compared to the LSTMs. The 152-layered ResNet and bidirectional LSTM showed an average false rejection rate of 8.34% and 16.67% and an equal error rate of 1.62% and 9.95%, respectively.
Highlights
The increasing adoption of biometrics in mobile devices has reduced the prevalence of traditional knowledge-based authentication methods, such as personal identification numbers (PINs), passwords, and pattern locks [1]
Based on the trained dataset, we evaluated the performance of residual networks (ResNets) (ResNet-50, ResNet-101, and ResNet 152) and two long short-term memory (LSTM) models with respect to age group classification and user authentication
This paper introduces the state-of-the-art convolutional neural network (CNN)-based ResNet models with a depth of up to 152 layers for user authentication in hand-object manipulation and compared its performance against the benchmark recurrent neural network (RNN)-based LSTMs
Summary
The increasing adoption of biometrics in mobile devices has reduced the prevalence of traditional knowledge-based authentication methods, such as personal identification numbers (PINs), passwords, and pattern locks [1]. Biometric authentication can be divided into two types: explicit and implicit [2]. Explicit biometric authentication uses physiological characteristics (e.g., fingerprints, irises, and face shapes) to verify the claimed identity of a user. Similar to knowledge-based authentication, explicit biometrics only authenticate a user at the initiation of a device or service, posing significant vulnerability to security attacks that might occur after the initial entry-point authentication [4,5]. Implicit authentication, which authenticates users based on behaviour patterns, enables implicit, continuous authentication as a background function in a device or service
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
