<![CDATA[The development of digital attendance systems requires authentication mechanisms that are not only flexible but also secure against identity and location manipulation. Conventional fingerprint-based attendance systems have limitations in supporting employee mobility and still require dedicated hardware at each workplace location. Existing online attendance systems generally implement GPS validation or face recognition independently and often lack mechanisms to prevent identity spoofing attacks. In addition, computationally intensive models may reduce system efficiency in real-time web environments. Therefore, this study proposes a web-based attendance system that integrates lightweight facial authentication, GPS-based geofencing validation, and passive liveness detection to improve both security and computational efficiency.This study aims to develop a web-based attendance system integrating ArcFace-based facial authentication, GPS-based geofencing validation, and Passive Liveness Detection using MiniFASNet. Experimental evaluations were conducted to assess authentication accuracy, computational efficiency, and spoofing detection capability. The results showed that MobileFaceNet achieved an authentication accuracy of 95.00% with an average inference time of 0.0216 seconds per face. In addition, the implemented liveness detection mechanism effectively detected most spoofing attempts involving photographs and replay videos. The system also achieved a False Acceptance Rate (FAR) of 0% under the evaluated experimental conditions and successfully fulfilled all functional requirements. Therefore, the proposed system provides a secure, efficient, and centralized attendance solution suitable for organizations with high employee mobility.]]>
