Faceless identification based on temporal strips
- PDF / 4,539,213 Bytes
- 20 Pages / 439.642 x 666.49 pts Page_size
- 23 Downloads / 186 Views
Faceless identification based on temporal strips Shu-Min Leong1
¨ C. -W. Phan1,2 · Vishnu Monn Baskaran1 · Chee-Pun Ooi3 · Raphael
Received: 17 September 2019 / Revised: 5 June 2020 / Accepted: 21 July 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper first presents a novel approach for modelling facial features, Local Directional Texture (LDT), which exploits the unique directional information in image textures for the problem of face recognition. A variant of LDT with privacy-preserving temporal strips (TS) is then considered to achieve faceless recognition with a higher degree of privacy while maintaining high accuracy. The TS uses two strips of pixel blocks from the temporal planes, XT and YT, for face recognition. By removing the reliance on spatial context (i.e., XY plane) for this task, the proposed method withholds facial appearance information from public view, where only one-dimensional temporal information that varies across time are extracted for recognition. Thus, privacy is assured, yet without impeding the facial recognition task which is vital for many security applications such as street surveillance and perimeter access control. To validate the reliability of the proposed method, experiments were carried out using the Honda/UCSD, CK+, CAS(ME)2 and CASME II databases. The proposed method achieved a recognition rate of 98.26% in the standard video-based face recognition database, Honda/UCSD. It also offers a 81.92% reduction in the dimension length required for storing the extracted features, in contrast to the conventional LBP-TOP. Keywords Facial recognition · Privacy-preserving · Temporal features · Local directional textures · Local binary pattern (LBP) · Partial facial features
Shu-Min Leong
[email protected] Rapha¨el C. -W. Phan [email protected] Vishnu Monn Baskaran [email protected] Chee-Pun Ooi [email protected] 1
School of Information Technology, Monash University, Subang Jaya, Malaysia
2
Department of Software Systems & Cybersecurity, Faculty of IT, Monash University, Melbourne, Australia
3
Faculty of Engineering, Multimedia University, Cyberjaya, Malaysia
Multimedia Tools and Applications
1 Introduction With the advancement in visual sensing capabilities of various IoT devices and real-world systems, face recognition has become a prevalent method to verify identities in order to control access to vital premises, critical resources, and even access to personal platforms enabled via mobile devices. A facial recognition system operates by verifying the person in front of it in comparison with stored static images or videos. Facial recognition systems have gone through significant improvements and are matured enough to be implemented in various sophisticated real-world environments that require high levels of accuracy. Although the technology started with recognizing faces from static images, research on using videos for recognition purposes has now gained significant attention in various applications such as surveillan
Data Loading...