EUVIMEDEUVIMED entdecken
Crossref

Dataset Distillation via a Noise-Unconstrained Generative Model

Jingxuan Zhang, Lei Dai, Fei Ye, Zhihua Chen, Ping Li, Xiaokang Yang, Bin Sheng

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-09 · 48 (9) · 10809-10826

DOI: 10.1109/tpami.2026.3690778 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-09. Verfasst von Jingxuan Zhang und 6 weiteren Autor:innen. Die Arbeit wurde bisher 27-mal zitiert (Crossref) und nennt 58 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Shenzhen Medical Research Fund, State Key Laboratory of Industrial Control Technology, China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dataset distillation (DD) aims to synthesize a more compact dataset than the original one and models trained on it are expected to have the same generalization capabilities as on the original dataset. Previous work via a generative model (GM) faces several limitations. First, GM struggles to generate representative samples due to a lack of constraints. Second, it overlooks the relationships between generated samples, limiting its effectiveness. In this paper, a new noise-unconstrained GM-based DD framework is proposed. In the distillation stage, an adaptive matching coefficient is introduced to align generated images with representative class elements and the MiniMax loss function is extended to reduce the optimization difficulty. In the deployment stage, features among each generative image are ensembled by gradient-matching based DD. Theoretical analysis based on McDiarmid's inequality demonstrates that the proposed components can reduce the generalization error of the original baseline method. We also provide insights into the potential of generated images as an effective proxy dataset for DD. For example, on the ImageWoof dataset with 50 distilled images per class using a 6-layer ConvNet for evaluation, generated images outperform 25%, 50%, and 75% original images by 8.4%, 6.3%, and 8.3% in distillation performance. Our method effectively handles both low- and high-resolution datasets, with experiments on 11 benchmarks demonstrating its efficacy.

Förderung

  • National Natural Science Foundation of China
  • Shenzhen Medical Research Fund
  • State Key Laboratory of Industrial Control Technology, China

Zitieren

APA

Zhang, J., Dai, L., Ye, F., Chen, Z., Li, P., Yang, X., & Sheng, B. (2026). Dataset Distillation via a Noise-Unconstrained Generative Model. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(9), 10809-10826. https://doi.org/10.1109/tpami.2026.3690778

Vancouver

Zhang J, Dai L, Ye F, Chen Z, Li P, Yang X, et al. Dataset Distillation via a Noise-Unconstrained Generative Model. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(9):10809-10826. doi:10.1109/tpami.2026.3690778

BibTeX
@article{Zhang2026,
  author = {Zhang, Jingxuan and Dai, Lei and Ye, Fei and Chen, Zhihua and Li, Ping and Yang, Xiaokang and Sheng, Bin},
  title = {Dataset Distillation via a Noise-Unconstrained Generative Model},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2026},
  volume = {48},
  number = {9},
  pages = {10809--10826},
  doi = {10.1109/tpami.2026.3690778}
}
Kostenlos starten →