High-Order Progressive Trajectory Matching for Medical Image Dataset Distillation
This addresses privacy and protocol barriers in medical image analysis by enabling more effective dataset distillation, though it appears incremental over existing trajectory matching approaches.
The paper tackles the challenge of medical image data sharing by proposing a high-order progressive trajectory matching method for dataset distillation, which improves distillation performance while maintaining model accuracy comparable to training on original datasets.
Medical image analysis faces significant challenges in data sharing due to privacy regulations and complex institutional protocols. Dataset distillation offers a solution to address these challenges by synthesizing compact datasets that capture essential information from real, large medical datasets. Trajectory matching has emerged as a promising methodology for dataset distillation; however, existing methods primarily focus on terminal states, overlooking crucial information in intermediate optimization states. We address this limitation by proposing a shape-wise potential that captures the geometric structure of parameter trajectories, and an easy-to-complex matching strategy that progressively addresses parameters based on their complexity. Experiments on medical image classification tasks demonstrate that our method improves distillation performance while preserving privacy and maintaining model accuracy comparable to training on the original datasets. Our code is available at https://github.com/Bian-jh/HoP-TM.