온라인 원격 수업 시스템에서 얼굴 프레젠테이션 공격 검출을 위한 모듈 구현 및 성능 향상

Abstract

Recently, due to situations such as Covid-19, various remote conferencing systems are coming to the market. As this situation is prolonged, online learning is becoming necessary in the educational environment. Face authentication for attendance is essential in the online class management system. However, there are various spoofing methods that attack these systems, and these include using a fake face through a printed photo or video playback. Therefore, the attendance system requires functions such as anti-spoofing to detect fake faces. In this paper, we design a presentation attack detection module and propose a method to effectively implement it.