A Contemporary Survey of Unimodal Liveness Detection Techniques: Challenges & Opportunities

Biometric recognition is an extensive and energizing field of research but spoofing attacks, "a deceiver pretense as an authorized user by reproducing the biometrics of the original user", is still the main problem of this field. Liveness detection is one prime countermeasure to spoofing attacks and is based on the principle that some additional information can be obtained or refined to verify that the produced data is authentic or not by a standard verification system. It utilized the physiological or anatomical signs of life, for example, change in, movement of the mouth, facial expression, blinking of eyes, etc. This paper presents an extensive analysis of the various unimodal biometric system based liveness/vitality detection techniques, in which physiological properties are utilized to discriminate between live and fake biometric traits. Further, these techniques are categorized based on static, dynamic, behavioral, and image quality based methods, recently proposed by various researchers. A new classification is also developed for these liveness detection techniques. The work covers theories, methodology, and evaluation datasets and also aims at future work in this field of research.