Enhancing Image Classification in Small and Unbalanced Datasets through Synthetic Data Augmentation

TL;DR

This paper introduces a synthetic data augmentation method using class-specific VAEs and latent space interpolation to improve image classification accuracy in small, imbalanced datasets, especially in medical imaging.

Contribution

The paper presents a novel augmentation strategy that generates realistic synthetic data via VAEs and latent space interpolation, enhancing model performance on small, imbalanced datasets.

Findings

Over 18% accuracy increase in underrepresented class

6% improvement in global accuracy and precision

Effective in small, imbalanced medical image datasets

Abstract

Accurate and robust medical image classification is a challenging task, especially in application domains where available annotated datasets are small and present high imbalance between target classes. Considering that data acquisition is not always feasible, especially for underrepresented classes, our approach introduces a novel synthetic augmentation strategy using class-specific Variational Autoencoders (VAEs) and latent space interpolation to improve discrimination capabilities. By generating realistic, varied synthetic data that fills feature space gaps, we address issues of data scarcity and class imbalance. The method presented in this paper relies on the interpolation of latent representations within each class, thus enriching the training set and improving the model's generalizability and diagnostic accuracy. The proposed strategy was tested in a small dataset of 321 images created to train and validate an automatic method for assessing the quality of cleanliness of esophagogastroduodenoscopy images. By combining real and synthetic data, an increase of over 18\% in the accuracy of the most challenging underrepresented class was observed. The proposed strategy not only benefited the underrepresented class but also led to a general improvement in other metrics, including a 6\% increase in global accuracy and precision.