ABSTRACT
The field of biometrics is actively exploring ear recognition as an ideal source for passive human identification. The human ear, being both unique and permanent, offers explicit advantages compared to other biometric options such as the retina, fingerprint, or face. In the paper, we have used YOLOv8 for ear detection for person recognition. 500 images each of a distinct individual are taken from two sources at Kaggle and split into a ratio of 7:2:1 for training, validation, and testing. The model has exhibited exemplary proficiency in person identification through ear biometrics, achieving an accuracy rate of 97.2%.
