Beyond Prototypes: Semantic Anchor Regularization for Better Representation Learning
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (3) · 1887-1895
DOI: 10.1609/aaai.v38i3.27958 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Yanqi Ge und 7 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref).
Abstract (Originalsprache)
One of the ultimate goals of representation learning is to achieve compactness within a class and well-separability between classes. Many outstanding metric-based and prototype-based methods following the Expectation-Maximization paradigm, have been proposed for this objective. However, they inevitably introduce biases into the learning process, particularly with long-tail distributed training data. In this paper, we reveal that the class prototype is not necessarily to be derived from training features and propose a novel perspective to use pre-defined class anchors serving as feature centroid to unidirectionally guide feature learning. However, the pre-defined anchors may have a large semantic distance from the pixel features, which prevents them from being directly applied. To address this issue and generate feature centroid independent from feature learning, a simple yet effective Semantic Anchor Regularization (SAR) is proposed. SAR ensures the inter-class separability of semantic anchors in the semantic space by employing a classifier-aware auxiliary cross-entropy loss during training via disentanglement learning. By pulling the learned features to these semantic anchors, several advantages can be attained: 1) the intra-class compactness and naturally inter-class separability, 2) induced bias or errors from feature learning can be avoided, and 3) robustness to the long-tailed problem. The proposed SAR can be used in a plug-and-play manner in the existing models. Extensive experiments demonstrate that the SAR performs better than previous sophisticated prototype-based methods. The implementation is available at https://github.com/geyanqi/SAR.
Zitieren
APA
Ge, Y., Nie, Q., Huang, Y., Liu, Y., Wang, C., Zheng, F., Li, W., & Duan, L. (2024). Beyond Prototypes: Semantic Anchor Regularization for Better Representation Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 38(3), 1887-1895. https://doi.org/10.1609/aaai.v38i3.27958
Vancouver
Ge Y, Nie Q, Huang Y, Liu Y, Wang C, Zheng F, et al. Beyond Prototypes: Semantic Anchor Regularization for Better Representation Learning. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(3):1887-1895. doi:10.1609/aaai.v38i3.27958
BibTeX
@article{Ge2024,
author = {Ge, Yanqi and Nie, Qiang and Huang, Ye and Liu, Yong and Wang, Chengjie and Zheng, Feng and Li, Wen and Duan, Lixin},
title = {Beyond Prototypes: Semantic Anchor Regularization for Better Representation Learning},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {3},
pages = {1887--1895},
doi = {10.1609/aaai.v38i3.27958}
}
EUVIMED