Saved in:
Bibliographic Details
Main Authors: Wang, Qichao, Lu, Yunhong, Cao, Hengyuan, Zhang, Junyi, Zhang, Min
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.03877
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917462273949696
author Wang, Qichao
Lu, Yunhong
Cao, Hengyuan
Zhang, Junyi
Zhang, Min
author_facet Wang, Qichao
Lu, Yunhong
Cao, Hengyuan
Zhang, Junyi
Zhang, Min
contents Dataset distillation enables efficient training by distilling the information of large-scale datasets into significantly smaller synthetic datasets. Diffusion based paradigms have emerged in recent years, offering novel perspectives for dataset distillation. However, they typically necessitate additional fine-tuning stages, and effective guidance mechanisms remain underexplored. To address these limitations, we rethink diffusion based dataset distillation and propose a Dual Matching Guided Diffusion (DMGD) framework, centered on efficient training-free guidance. We first establish Semantic Matching via conditional likelihood optimization, eliminating the need for auxiliary classifiers. Furthermore, we propose a dynamic guidance mechanism that enhances the diversity of synthetic data while maintaining semantic alignment. Simultaneously, we introduce an optimal transport (OT) based Distribution Matching approach to further align with the target distribution structure. To ensure efficiency, we develop two enhanced strategies for diffusion based framework: Distribution Approximate Matching and Greedy Progressive Matching. These strategies enable effective distribution matching guidance with minimal computational overhead. Experimental results on ImageNet-Woof, ImageNet-Nette, and ImageNet-1K demonstrate that our training-free approach achieves significant improvements, outperforming state-of-the-art (SOTA) methods requiring additional fine-tuning by average accuracy gains of 2.1%, 5.4%, and 2.4%, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2605_03877
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle DMGD: Train-Free Dataset Distillation with Semantic-Distribution Matching in Diffusion Models
Wang, Qichao
Lu, Yunhong
Cao, Hengyuan
Zhang, Junyi
Zhang, Min
Computer Vision and Pattern Recognition
Artificial Intelligence
Dataset distillation enables efficient training by distilling the information of large-scale datasets into significantly smaller synthetic datasets. Diffusion based paradigms have emerged in recent years, offering novel perspectives for dataset distillation. However, they typically necessitate additional fine-tuning stages, and effective guidance mechanisms remain underexplored. To address these limitations, we rethink diffusion based dataset distillation and propose a Dual Matching Guided Diffusion (DMGD) framework, centered on efficient training-free guidance. We first establish Semantic Matching via conditional likelihood optimization, eliminating the need for auxiliary classifiers. Furthermore, we propose a dynamic guidance mechanism that enhances the diversity of synthetic data while maintaining semantic alignment. Simultaneously, we introduce an optimal transport (OT) based Distribution Matching approach to further align with the target distribution structure. To ensure efficiency, we develop two enhanced strategies for diffusion based framework: Distribution Approximate Matching and Greedy Progressive Matching. These strategies enable effective distribution matching guidance with minimal computational overhead. Experimental results on ImageNet-Woof, ImageNet-Nette, and ImageNet-1K demonstrate that our training-free approach achieves significant improvements, outperforming state-of-the-art (SOTA) methods requiring additional fine-tuning by average accuracy gains of 2.1%, 5.4%, and 2.4%, respectively.
title DMGD: Train-Free Dataset Distillation with Semantic-Distribution Matching in Diffusion Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2605.03877