Large Variability Surveillance Camera Face Database
Tanko Daniel Salka, Marsyita Binti Hanafi, Syamsiah Mashohor, Sharifah Mumtazah Syed Ahamad · 2015
Recognizing and verifying human faces captured by surveillance camera are critically important and challenging tasks. For these purposes, many databases have been introduced but most of them have less variability and small number of subjects. In this paper, we introduce a large scale, more challenging and large variability surveillance camera face database, which cater the need of tackling the problems related to face recognition and verification. The database consists of 20,000 still images and 200 videos of 100 subjects. The images are captured under controlled and uncontrolled conditions with consideration of day and night modes. The database also exhibits the operational variability in face images such as pose, expressions and occlusion and it would provide a more challenging and realistic scenarios for the studies of face detection and recognition.