Class-Center Involved Triplet Loss for Skin Disease Classification on Imbalanced Data
Weixian Lei, Rong Zhang, Yang Yang, Ruixuan Wang, Wei‐Shi Zheng · 2020
It is ideal to develop intelligent systems to accurately diagnose diseases as human specialists do. However, due to the highly imbalanced data issue between common and rare diseases, it is still an open problem for the systems to effectively learn to recognize both common and rare diseases. In this paper, we novelly applied triplet modelling to overcome the data imbalance issue particularly for diagnosis of rare diseases. Moreover, we further applied a class-center based triplet loss in order to make the triplet-based learning more stable. Extensive evaluation on two skin image classification tasks shows that the triplet-based approach is very effective and outperforms the widely used methods for solving the imbalance problem.