Steering Diffusion Models via Class-Contrastive Influence for Few-Shot Medical Classification

2026-07-14

AI Summary

In few-shot medical image classification, synthetic samples generated by diffusion models often exhibit limited utility for downstream tasks and lack targeted optimization. To address this, this work proposes the Class-Contrastive Influence (C2I) criterion, which introduces gradient-based influence analysis into synthetic data generation for the first time. By quantifying each synthetic sample’s contribution to the classifier’s decision boundary, C2I constructs a reinforcement learning reward signal to fine-tune the diffusion model, steering it to generate samples that are more discriminative and lie near class boundaries. This approach departs from conventional paradigms that prioritize photorealism or diversity alone, achieving significant performance gains over existing diffusion-based data augmentation methods across multiple few-shot medical imaging benchmarks, thereby enhancing both classification accuracy and model robustness.

Abstract

When labeled data are scarce, off-the-shelf diffusion models can augment training sets for few-shot medical image classification, but not all generated samples are equally useful for the downstream task. Existing approaches largely improve synthetic data by increasing realism, diversity, or domain adaptation, while overlooking a more fundamental question: how should sample usefulness for classification be measured and optimized? We address this with Class-Contrastive Influence (C2I), a criterion that quantifies a sample's usefulness through its gradient-based influence on the classifier. We find that effective samples exhibit a strong C2I gap: their loss gradients align with validation gradients from the same class and oppose those from other classes. Our analysis further suggests that such high-C2I samples are hard, boundary-proximal examples that help refine the decision boundary and improve robustness. Building on this insight, we fine-tune diffusion models with reinforcement learning using a C2I-based reward to steer generation toward class-informative samples. Across several few-shot medical imaging benchmarks, C2I-guided generation improves downstream accuracy and robustness over diffusion-based augmentation baselines, showing that synthetic augmentation is most effective when guided by task usefulness rather than image quality alone.

Problem

• few-shot medical classification
• diffusion models
• sample usefulness
• class-contrastive influence
• synthetic data augmentation

Innovation

• Class-Contrastive Influence
• Diffusion Models
• Few-Shot Medical Classification
• Task-Guided Generation
• Gradient-Based Influence