EUVIMEDEUVIMED entdecken
Crossref

Exploring Diverse Representations for Open Set Recognition

Yu Wang, Junxian Mu, Pengfei Zhu, Qinghua Hu

Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (6) · 5731-5739

DOI: 10.1609/aaai.v38i6.28385 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Yu Wang und 3 weiteren Autor:innen. Die Arbeit wurde bisher 31-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Open set recognition (OSR) requires the model to classify samples that belong to closed sets while rejecting unknown samples during test. Currently, generative models often perform better than discriminative models in OSR, but recent studies show that generative models may be computationally infeasible or unstable on complex tasks. In this paper, we provide insights into OSR and find that learning supplementary representations can theoretically reduce the open space risk. Based on the analysis, we propose a new model, namely Multi-Expert Diverse Attention Fusion (MEDAF), that learns diverse representations in a discriminative way. MEDAF consists of multiple experts that are learned with an attention diversity regularization term to ensure the attention maps are mutually different. The logits learned by each expert are adaptively fused and used to identify the unknowns through the score function. We show that the differences in attention maps can lead to diverse representations so that the fused representations can well handle the open space. Extensive experiments are conducted on standard and OSR large-scale benchmarks. Results show that the proposed discriminative method can outperform existing generative models by up to 9.5% on AUROC and achieve new state-of-the-art performance with little computational cost. Our method can also seamlessly integrate existing classification models. Code is available at https://github.com/Vanixxz/MEDAF.

Zitieren

APA

Wang, Y., Mu, J., Zhu, P., & Hu, Q. (2024). Exploring Diverse Representations for Open Set Recognition. Proceedings of the AAAI Conference on Artificial Intelligence, 38(6), 5731-5739. https://doi.org/10.1609/aaai.v38i6.28385

Vancouver

Wang Y, Mu J, Zhu P, Hu Q. Exploring Diverse Representations for Open Set Recognition. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(6):5731-5739. doi:10.1609/aaai.v38i6.28385

BibTeX
@article{Wang2024,
  author = {Wang, Yu and Mu, Junxian and Zhu, Pengfei and Hu, Qinghua},
  title = {Exploring Diverse Representations for Open Set Recognition},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2024},
  volume = {38},
  number = {6},
  pages = {5731--5739},
  doi = {10.1609/aaai.v38i6.28385}
}
Kostenlos starten →