EUVIMEDEUVIMED entdecken
Crossref

Filter Pruning by High-Order Spectral Clustering

Hang Lin, Yifan Peng, Yubo Zhang, Lin Bie, Xibin Zhao, Yue Gao

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2025-04 · 47 (4) · 2402-2415

DOI: 10.1109/tpami.2024.3524381 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025-04. Verfasst von Hang Lin und 5 weiteren Autor:innen. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 66 Literaturangaben. Gefördert von: NSFC Program, Industrial Technology Infrastructure Public Service Platform Project through the Public Service Platform for Urban Rail Transit Equipment Signal System Testing…, Ministry of Industry and Information Technology of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Large amount of redundancy is widely present in convolutional neural networks (CNNs). Identifying the redundancy in the network and removing the redundant filters is an effective way to compress the CNN model size with a minimal reduction in performance. However, most of the existing redundancy-based pruning methods only consider the distance information between two filters, which can only model simple correlations between filters. Moreover, we point out that distance-based pruning methods are not applicable for high-dimensional features in CNN models by our experimental observations and analysis. To tackle this issue, we propose a new pruning strategy based on high-order spectral clustering. In this approach, we use hypergraph structure to construct complex correlations among filters, and obtain high-order information among filters by hypergraph structure learning. Finally, based on the high-order information, we can perform better clustering on the filters and remove the redundant filters in each cluster. Experiments on various CNN models and datasets demonstrate that our proposed method outperforms the recent state-of-the-art works. For example, with ResNet50, we achieve a 57.1% FLOPs reduction with no accuracy drop on ImageNet, which is the first to achieve lossless pruning with such a high compression ratio.

Förderung

  • NSFC Program
  • Industrial Technology Infrastructure Public Service Platform Project through the Public Service Platform for Urban Rail Transit Equipment Signal System Testing…
  • Ministry of Industry and Information Technology of China
  • Science and Technology Innovation Project of Hunan Province
  • Beijing Natural Science Foundation
  • National Key Laboratory of Human-Machine Hybrid Augmented Intelligence, Xi’an Jiaotong University

Zitieren

APA

Lin, H., Peng, Y., Zhang, Y., Bie, L., Zhao, X., & Gao, Y. (2025). Filter Pruning by High-Order Spectral Clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(4), 2402-2415. https://doi.org/10.1109/tpami.2024.3524381

Vancouver

Lin H, Peng Y, Zhang Y, Bie L, Zhao X, Gao Y. Filter Pruning by High-Order Spectral Clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2025;47(4):2402-2415. doi:10.1109/tpami.2024.3524381

BibTeX
@article{Lin2025,
  author = {Lin, Hang and Peng, Yifan and Zhang, Yubo and Bie, Lin and Zhao, Xibin and Gao, Yue},
  title = {Filter Pruning by High-Order Spectral Clustering},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2025},
  volume = {47},
  number = {4},
  pages = {2402--2415},
  doi = {10.1109/tpami.2024.3524381}
}
Kostenlos starten →