Comparison of Vision Transformers and InceptionV3 on Multiclass Skin Disease Image Classification

Abstract

Skin diseases pose a major health problem to the world population necessitating proper and timely diagnostic modalities to improve patient care and clinical outcomes. This paper presents a comparative analysis of Vision Transformers (ViT) and InceptionV3 for multiclass skin disease image classification across 22 different dermatological conditions. Empirical evidence shows that Vision Transformers consistently achieve higher overall accuracy and class-specific discriminative performance than InceptionV3.

Keywords

Disease Classification, Vision Transformer, InceptionV3, Deep Learning, Medical Image Analysis, Dermatology, Multi-Class Classification

Download Article

File: Shruthisha-S-Nair.pdf | Size: N/A

Citation