Implementasi Metode Object Detection Dengan Algoritma You Only Look Once (YOLO) Untuk Deteksi Kecurangan Di Dalam Ruang Ujian

Sari

Cheating behavior is a dishonest act carried out by someone to get a satisfactory end result. Academic cheating is still often done by students to get high grades. In recent years, research on artificial intelligence such as object detection has been carried out and the results can make it easier for researchers to recognize objects contained in an image. Therefore, in this study, a system will be built that is able to detect the experience of examinees in the room during the exam. Detection is done using a camera installed in the exam room, the camera will take real-time video images that will be used as input. In this study the author uses yahoo You Only Look Once (YOLO) version 4. This study uses a dataset of 1050 images divided into 6 classes, namely: students using cellphones, cellphones, looking left, looking right, looking down, looking back, and students who work together. The results show that the YOLOv4 algorithm can recognize and detect objects in real time using self-trained weights with a Mean Average Precision (mAP) of 86%.

Keywords—YOLO, Cheating, Object Detection, Computer Vision, exam, Digital Image