A Computer Vision-Based System for Monitoring Liveness in Remote Work

Filipe Oliveira De Saldanha Da Gama, Rafael Senra Donner Jorge, Caio Rangel Fernandes, Marcio Moreira Do Nascimento Filho, Rigel Procópio Fernandes, Thiago Silva de Souza

DOI: 10.14209/sbrt.2025.1571143918

Evento: XLIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2025)

Keywords:

Abstract

This undergraduate paper proposes a home office time clock system that uses images captured from the user's webcam to detect the liveness of employees. Growing concerns about productivity and accountability in remote work environments have created a strong demand for effective monitoring solutions. This work evaluates computer vision techniques for tracking employee activity in such settings, offering a non-intrusive and reliable system suitable for real-world deployment. Experimental results from tests performed under challenging conditions, including low-light environments, demonstrate that the system achieves a 100% detection rate in captured frames at ambient illumination levels near 7 lux.

Download