Expediting Automated Face Recognition Using the Novel ORB2-IPR Framework
Face Recognition (FR) is at the forefront of distinctly unresolved challenges in the domain of Computer Vision, due to the sharp accuracy and performance drops it undergoes, when there are pronounced variations in parameters such as illumination, pose, ba
- PDF / 253,214 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 79 Downloads / 221 Views
Abstract Face Recognition (FR) is at the forefront of distinctly unresolved challenges in the domain of Computer Vision, due to the sharp accuracy and performance drops it undergoes, when there are pronounced variations in parameters such as illumination, pose, background clutter and so on between the input and database faces. In this paper, we attempt to expedite the performance of automated FR with real-time images, using a novel framework called ORB2-IPR (ORB based Bag of Interest Points using RANSAC), which exhaustively learns a vocabulary of highly discriminative facial interest points from the facial database images (which can be referred to, and compared directly, instead of following the conventional time-intensive approach of comparing a given input face with each database face separately) by employing the cost-effective ORB (Oriented Fast Rotated Brief) descriptor (instead of the commonly employed SIFT and SURF descriptors), followed by the application of RANSAC (Random Sample Consensus) as a post-processing step to remove noise in the form of outliers, in order to improve the accuracy of the system. We will conclusively demonstrate that our technique is capable of rendering superior performance than the state-of-the-art methodologies using extensive mathematical arguments and by carrying out ample experimentations on the benchmark ORL, Face 95 and LFW databases.
A. Vinay (&) V.S. Shekhar N. Manjunath K.N. Balasubramanya Murthy S. Natarajan PES University, 100 Feet Ring Road, BSK 3rd Stage, Bengaluru 560085, India e-mail: [email protected] V.S. Shekhar e-mail: [email protected] N. Manjunath e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2018 D.S. Guru et al. (eds.), Proceedings of International Conference on Cognition and Recognition, Lecture Notes in Networks and Systems 14, DOI 10.1007/978-981-10-5146-3_21
223
224
A. Vinay et al.
1 Introduction Face recognition (FR) [1] is one of the predominant forms of Biometrics whose applications are apparent in a multitude of commercial and law-enforcement settings. Even though FR finds implementation in a plethora of mainstream applications, its performance is severely compromised in unconstrained scenarios i.e. scenarios involving sharp variations between the input and database faces with respect to factors such as illumination, background clutter, Occlusion (hindrances such as Beard, Glasses etc.) affine (camera angle during image acquisition) [2] and so on. Furthermore, an efficient FR system is heavily reliant on the selection of the Feature Extractor to be employed; to that end, many state-of-the-art algorithms such as SIFT [3, 4, 33], SURF [5, 6], BRISK [7], BRIEF [8] are available (a thorough overview evaluation of the popular Feature Extraction algorithms is provided in [9]). Although, these descriptors work competently in most of the constrained and some unconstrained scenarios, these algorithms demand substantial computational power, and hence their implementation for real-time settings (which most definitely will inclu
Data Loading...
