Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.11434 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915812592320512 |
|---|---|
| author | Ding, Yifan Liu, Xixi Unger, Jonas Eilertsen, Gabriel |
| author_facet | Ding, Yifan Liu, Xixi Unger, Jonas Eilertsen, Gabriel |
| contents | \noindent Out-of-distribution (OOD) detection is essential for the safe deployment of machine learning models. Extensive work has focused on devising various scoring functions for detecting OOD samples, while only a few studies focus on training neural networks using certain model calibration objectives, which often lead to a compromise in predictive accuracy and support only limited choices of scoring functions. In this work, we first identify the feature collapse phenomena in Logit Normalization (LogitNorm), then propose a novel hyperparameter-free formulation that significantly benefits a wide range of post-hoc detection methods. To be specific, we devise a feature distance-awareness loss term in addition to LogitNorm, termed $\textbf{ELogitNorm}$, which enables improved OOD detection and in-distribution (ID) confidence calibration. Extensive experiments across standard benchmarks demonstrate that our approach outperforms state-of-the-art training-time methods in OOD detection while maintaining strong ID classification accuracy. Our code is available on: https://github.com/limchaos/ElogitNorm. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_11434 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Out-of-Distribution Detection with Extended Logit Normalization Ding, Yifan Liu, Xixi Unger, Jonas Eilertsen, Gabriel Computer Vision and Pattern Recognition \noindent Out-of-distribution (OOD) detection is essential for the safe deployment of machine learning models. Extensive work has focused on devising various scoring functions for detecting OOD samples, while only a few studies focus on training neural networks using certain model calibration objectives, which often lead to a compromise in predictive accuracy and support only limited choices of scoring functions. In this work, we first identify the feature collapse phenomena in Logit Normalization (LogitNorm), then propose a novel hyperparameter-free formulation that significantly benefits a wide range of post-hoc detection methods. To be specific, we devise a feature distance-awareness loss term in addition to LogitNorm, termed $\textbf{ELogitNorm}$, which enables improved OOD detection and in-distribution (ID) confidence calibration. Extensive experiments across standard benchmarks demonstrate that our approach outperforms state-of-the-art training-time methods in OOD detection while maintaining strong ID classification accuracy. Our code is available on: https://github.com/limchaos/ElogitNorm. |
| title | Enhancing Out-of-Distribution Detection with Extended Logit Normalization |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.11434 |