Automatic Face Detection Uses Deep Learning to Prevent Cheating in General Election
Labib Yusuf Aditama, Uus Khusni, Sriyadi, Sulistiyah, Ivan Aprizal, Meryl Putra Pratama · 2024
General elections are the cornerstone of a democratic system, where legitimacy and public trust are crucial for maintaining political stability and social welfare. In this context, challenges such as systematic fraud and the complexity of voter identification are primary concerns. This study analyzes the effectiveness of visual identity detection technology in ensuring electoral integrity and reducing fraud risk. Utilizing Convolutional Neural Networks (CNNs) as a part of Deep Learning (DL) is expected to address this issue. This study evaluated the visual identity detection system on a dataset of faces with varying perspectives and accessories. The test results showed that the model accurately recognized the faces, achieving an accuracy rate of approximately 90%. The analysis concludes that advanced research is needed for further development with more diverse datasets to enhance the accuracy in various complex situations. This study provides insights into the application of visual identity detection technology in general elections, aiming to contribute to strengthening public trust and security in the democratic process.