Supervised contrastive pretrained ResNet with MixUp to enhance respiratory sound classification on imbalanced and limited dataset

Jinhai Hu, Cong Sheng Leow, Shuailin Tao, Wang Ling Goh, Yuan Gao

2023 IEEE Biomedical Circuits and Systems Conference (BioCAS 2023), 2023

Contrastive Learning · Limited and Imbalanced Data

Overview

A classification approach combining supervised contrastive pretraining, feature augmentation, and MixUp to learn from limited and imbalanced data. The work evaluates these techniques on respiratory sound classification tasks.