%0 Conference Paper
%A Shen, Ng Yong
%A Yi, Kh'Ng Xin
%A Lin, Beh Woan
%A Ooi, Boonyaik Yaik
%D 2025
%F scholars:20414
%I Institute of Electrical and Electronics Engineers Inc.
%K Convolutional neural networks; Deep learning; Learning algorithms; Learning systems; Accurate analysis; Comparative analyzes; Convolutional neural network; Efficientnetb0; GPT assistant-based; Intelligence models; Resnet50; Skin analysis; Skin conditions; YOLOv8; Classification (of information)
%P 455 - 460
%R 10.1109/AiDAS67696.2025.11213898
%T Comparative Analysis of AI Models for Skin Type and Acne Severity Classification
%U https://khub.utp.edu.my/scholars/20414/
%X Skin acne is a common problem faced by many people worldwide, which will affect both confidence and physical appearance of individuals. Accurate analysis of skin condition is essential to ensure that users can understand their skin condition. Recently, artificial intelligence (AI) has demonstrated strong potential in overcoming these challenges by delivering accurate results. In the field of AI, techniques like machine learning and deep learning are increasingly applied to the analysis of skin conditions. While deep learning has shown strong potential, its accuracy remains limited in certain scenarios. To address this, the project aims to identify the most effective skin analysis model among Classic CNN, ResNet50, EfficientNetB0, and YOLOv8. These models are evaluated based on their ability to accurately assess users' skin type and acne severity. Furthermore, the integration of GPT-4o's natural language processing capabilities introduces a novel GPT assistant-based approach. This complements traditional deep learning methods by enhancing interpretability and flexibility in prediction tasks, thereby offering a more comprehensive and user-friendly skin analysis solution. © 2025 IEEE.
%Z Cited by: 0