EUVIMEDEUVIMED entdecken
Crossref

Continual Out-of-Distribution Detection with Analytic Neural Collapse

Saleh Momeni, Changnan Xiao, Bing Liu

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (18) · 15564-15572

DOI: 10.1609/aaai.v40i18.38585 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Saleh Momeni und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Continual learning (CL) aims to enable models to incrementally learn from a sequence of tasks without forgetting previously acquired knowledge. While most prior work focuses on closed-world settings, where all test instances are assumed from the set of learned classes, real-world applications require models to handle both CL and out-of-distribution (OOD) samples. A key insight from recent studies on deep neural networks is the phenomenon of Neural Collapse (NC), which occurs in the terminal phase of training when the loss approaches zero. Under NC, class features collapse to their means, and classifier weights align with these means, enabling effective prototype-based strategies such as nearest class mean, for both classification and OOD detection. However, in CL, catastrophic forgetting (CF) prevents the model from naturally reaching this desirable regime. In this paper, we propose a novel method called Analytic Neural Collapse (AnaNC) that analytically creates the NC properties in the feature space of a frozen pre-trained model with no training, overcoming CF. Extensive experiments demonstrate that our approach outperforms state-of-the-art methods in continual OOD detection and learning, highlighting the effectiveness of our method in this challenging scenario.

Zitieren

APA

Momeni, S., Xiao, C., & Liu, B. (2026). Continual Out-of-Distribution Detection with Analytic Neural Collapse. Proceedings of the AAAI Conference on Artificial Intelligence, 40(18), 15564-15572. https://doi.org/10.1609/aaai.v40i18.38585

Vancouver

Momeni S, Xiao C, Liu B. Continual Out-of-Distribution Detection with Analytic Neural Collapse. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(18):15564-15572. doi:10.1609/aaai.v40i18.38585

BibTeX
@article{Momeni2026,
  author = {Momeni, Saleh and Xiao, Changnan and Liu, Bing},
  title = {Continual Out-of-Distribution Detection with Analytic Neural Collapse},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {18},
  pages = {15564--15572},
  doi = {10.1609/aaai.v40i18.38585}
}
Kostenlos starten →