Multi-Category Fusion Contrastive Learning with Core Data Selection for Robust RGB Image-based Dental Caries Classification

Peiliang Zhang, Yaru Chen, Yunjiong Liu, Chao Che, Yongjun Zhu

Abstract: Highlights•M3C is a model structure-independent dental caries detection pipeline.•M3C robustness verified from model representation, theoretical analysis and mutual information computation.•Core data selection and inter-category contrastive learning improve M3C robustness.•M3C outperforms the SOTA dental caries detection model using only 50% of core data.