Acne Type Recognition for Mobile-Based Application Using YOLO

• Published in: Journal of Physics: Conference Series
• Main Author: Isa N.A.M.; Mangshor N.N.A.
• Format: Conference paper
• Language: English
• Published: IOP Publishing Ltd 2021
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85111970547&doi=10.1088%2f1742-6596%2f1962%2f1%2f012041&partnerID=40&md5=c764f6d18c42db148f5781a62b5dc3c3

Acne is a chronic skin condition that happens to most teenagers at the age of 12 and 25. Several types of acne are found as non-inflammatory and inflammatory skin disease. As the number of people facing acne problems increasing and there is the need to have an automated application for recognizing acne, this study proposed a mobile-based application that able to recognizes acne types. This study used the Deep Learning technology method, YOLOv4 in detecting and recognizing acne. There are four types of acne covered in this study which are cyst, papule, pustule, and whitehead. The dataset used for the purpose of training and testing the model is from the DermNet NZ dataset. Based on the testing conducted, the application achieved 91.25% average of accuracy. It is believe, the integration of YOLO with any existing Deep Learning approach could improve the recognition rate in future. © Published under licence by IOP Publishing Ltd.