Ear detection based on YOLOv3

Abstract

Human ear has received loads of attentions recently due to its distinctiveness. Therefore, many ear-based applications has been proposed with significant perfomance. However, the lack of efficient ear detection methods is remaining. In this paper, we present a new ear detector based on YOLOv3. YOLO is now one of the fastest detectors that is widely applied in many real-time applications. A face database is employed for evaluating.