Dataset Condensation Atlas

Method · Optimization and training recipes

Not All Samples Should Be Utilized Equally: Towards Understanding and Improving Dataset Distillation

Shaobo Wang, Yantai Yang, Qilong Wang, Kaixin Li, Linfeng Zhang, Junchi Yan

CVPR 2025 Workshop · first public 2024-08-22 · arXiv 2408.12483

paper ↗catalogued✓ abstract read

In one paragraph

Measures sample difficulty by gradient norm and finds that different matching-based distillation methods implicitly favor samples of different difficulty, then extends neural scaling laws for data pruning to dataset distillation to argue that prioritizing easier real samples during synthesis improves the distilled set, especially at low IPC, packaging this as a Sample Difficulty Correction plug-in that improves 7 distillation methods across 6 datasets.

Where it sits

Abstract (verbatim from arXiv)

Dataset Distillation (DD) aims to synthesize a small dataset capable of performing comparably to the original dataset. Despite the success of numerous DD methods, theoretical exploration of this area remains unaddressed. In this paper, we take an initial step towards understanding various matching-based DD methods from the perspective of sample difficulty. We begin by empirically examining sample difficulty, measured by gradient norm, and observe that different matching-based methods roughly correspond to specific difficulty tendencies. We then extend the neural scaling laws of data pruning to DD to theoretically explain these matching-based methods. Our findings suggest that prioritizing the synthesis of easier samples from the original dataset can enhance the quality of distilled datasets, especially in low IPC (image-per-class) settings. Based on our empirical observations and theoretical analysis, we introduce the Sample Difficulty Correction (SDC) approach, designed to predominantly generate easier samples to achieve higher dataset quality. Our SDC can be seamlessly integrated into existing methods as a plugin with minimal code adjustments. Experimental results demonstrate that adding SDC generates higher-quality distilled datasets across 7 distillation methods and 6 datasets.

BibTeX (generated; prefer the venue's official entry)
@article{wang2024samples,
  title   = {Not All Samples Should Be Utilized Equally: Towards Understanding and Improving Dataset Distillation},
  author  = {Shaobo Wang and Yantai Yang and Qilong Wang and Kaixin Li and Linfeng Zhang and Junchi Yan},
  journal = {CVPR 2025 Workshop},
  year    = {2024}
}

Nearby in Optimization and training recipes

2026-05

C^2R — Mind Your Margin and Boundary: Are Your Distilled Datasets Truly Robust?

Muquan Li, Yingyi Ma, Yihong Huang et al. · ICML 2026notablepaper ↗

2026-04

COBRA — Fair Dataset Distillation via Cross-Group Barycenter Alignment

Mohammad Hossein Moslemi, Nima Hosseini Dashtbayaz, Zhimin Mei et al. · ICML 2026notablepaper ↗code ↗

2026-03

PTM-ST — Multimodal Dataset Distillation via Phased Teacher Models

Shengbin Guo, Hang Zhao, Senqiao Yang et al. · ICLR 2026notableVision–languagepaper ↗code ↗

2026-03

FD2 — FD$^2$: A Dedicated Framework for Fine-Grained Dataset Distillation

Hongxu Ma, Guang Li, Shijie Wang et al. · ECCV 2026notablepaper ↗

2025-05

PRISM — PRISM: Video Dataset Condensation with Progressive Refinement and Insertion for Sparse Motion

Jaehyun Choi, Jiwan Hur, Gyojin Han et al. · CVPR 2026notableVideopaper ↗