Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Iterative Online Image Synthesis via Diffusion Model for Imbalanced Classification

Shuhan Li, Yi Lin, Hao Chen, Kwang-Ting Cheng

TL;DR

An Iterative Online Image Synthesis (IOIS) framework to address the class imbalance problem in medical image classification is introduced and the results obtained demonstrate the superiority of the approach over state-of-the-art methods as well as the effectiveness of each component.

Abstract

Accurate and robust classification of diseases is important for proper diagnosis and treatment. However, medical datasets often face challenges related to limited sample sizes and inherent imbalanced distributions, due to difficulties in data collection and variations in disease prevalence across different types. In this paper, we introduce an Iterative Online Image Synthesis (IOIS) framework to address the class imbalance problem in medical image classification. Our framework incorporates two key modules, namely Online Image Synthesis (OIS) and Accuracy Adaptive Sampling (AAS), which collectively target the imbalance classification issue at both the instance level and the class level. The OIS module alleviates the data insufficiency problem by generating representative samples tailored for online training of the classifier. On the other hand, the AAS module dynamically balances the synthesized samples among various classes, targeting those with low training accuracy. To evaluate the effectiveness of our proposed method in addressing imbalanced classification, we conduct experiments on the HAM10000 and APTOS datasets. The results obtained demonstrate the superiority of our approach over state-of-the-art methods as well as the effectiveness of each component. The source code will be released upon acceptance.

Iterative Online Image Synthesis via Diffusion Model for Imbalanced Classification

TL;DR

An Iterative Online Image Synthesis (IOIS) framework to address the class imbalance problem in medical image classification is introduced and the results obtained demonstrate the superiority of the approach over state-of-the-art methods as well as the effectiveness of each component.

Abstract

Accurate and robust classification of diseases is important for proper diagnosis and treatment. However, medical datasets often face challenges related to limited sample sizes and inherent imbalanced distributions, due to difficulties in data collection and variations in disease prevalence across different types. In this paper, we introduce an Iterative Online Image Synthesis (IOIS) framework to address the class imbalance problem in medical image classification. Our framework incorporates two key modules, namely Online Image Synthesis (OIS) and Accuracy Adaptive Sampling (AAS), which collectively target the imbalance classification issue at both the instance level and the class level. The OIS module alleviates the data insufficiency problem by generating representative samples tailored for online training of the classifier. On the other hand, the AAS module dynamically balances the synthesized samples among various classes, targeting those with low training accuracy. To evaluate the effectiveness of our proposed method in addressing imbalanced classification, we conduct experiments on the HAM10000 and APTOS datasets. The results obtained demonstrate the superiority of our approach over state-of-the-art methods as well as the effectiveness of each component. The source code will be released upon acceptance.
Paper Structure (14 sections, 5 equations, 3 figures, 1 table, 1 algorithm)

This paper contains 14 sections, 5 equations, 3 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Methodology
  3. Diffusion Model
  4. Online Image Synthesis with Classifier Guidance
  5. Accuracy Adaptive Sampling for Imbalanced Data
  6. Experiments
  7. Datasets and Evaluation Metrics
  8. Impelementation Details
  9. Comparisons with State-of-the-art Methods
  10. Ablation Study
  11. Discussions
  12. Class-wise comparison.
  13. Visualization of synthetic images.
  14. Conclusion

Figures (3)

  • Figure 1: The framework of the proposed IOIS. It comprises three conponents, which are iteratively performed: training classifier, determining class distribution of synthetic images via AAS, and synthesizing images via OIS.
  • Figure 2: Comparisons of per class accuracy of different methods on HAM10000 dataset. The relative variations of each class accuracy compared to the CE loss are computed. The numbers in brackets following the class name denote the size of samples in that class.
  • Figure 3: Visualization examples of synthesized images and real images.