Abstract
Three-dimensional (3D) face models can intrinsically handle large pose face recognition problem. In this paper, we propose a novel pose-invariant face recognition method via RGB-D images. By employing depth, our method is able to handle self-occlusion and deformation, both of which are challenging problems in two-dimensional (2D) face recognition. Texture images in the gallery can be rendered to the same view as the probe via depth. Meanwhile, depth is also used for similarity measure via frontalization and symmetric filling. Finally, both texture and depth contribute to the final identity estimation. Experiments on Bosphorus, CurtinFaces, Eurecom, and Kiwi databases demonstrate that the additional depth information has improved the performance of face recognition with large pose variations and under even more challenging conditions.
Highlights
Face recognition has been attracting considerable attention from researchers due to its wide variety of applications, such as homeland security, video surveillance, law enforcement, and identity management
One frontal neutral face model is used as the gallery set
All the frontal view face models for each subject are used to train 13 texture and depth Joint Bayesian classifiers. 13 images with yaw, pitch, and cross rotation incorporate both yaw and pitch for each subject are used as the probe set
Summary
Face recognition has been attracting considerable attention from researchers due to its wide variety of applications, such as homeland security, video surveillance, law enforcement, and identity management. Performance of face recognition is improving [1,2,3,4,5], but large pose variation problem still remains unsolved [6,7,8]. Reference [9] reconstructed a 3D shape model from multiple 2D face images and generated a set of densely sampled 2D face images as templates for pose-invariant recognition. Reference [10] proposed a fully automatic system that matched the reconstructed faces under frontal view. It can handle continuous pose variation up to ±45 in yaw and ±30 in pitch angles. Precise facial landmark still remains unsolved under large pose variations
Sign-in/Register to view highlights & summary 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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
