Eye Pupil Localisation and Labeling Using a Small Size Database and YOLOv4 Object Detection Algorithm

Eye-related research has shown that eye gaze data are very important for applications essential to human daily life. Eye been used in and systems movements, tracking tracking. pupil localization, labelling challenging problems computer science. This article discusses explores problem․ The YOLOv4 (“You only look once”) object detection algorithm which is an evolution of the YOLOv3 model evaluated a tiny database consisting 103 images. was created by Alexey Bochkovskiy, Chien-Yao Wang Hong-Yuan Mark Liao [3]. It twice as fast EfficientDet with comparable performance. main purpose this test algorithm, find out its effectiveness process localization pupils (determine) when training relatively small number iterations.