Frustratingly Easy Feature Reconstruction for Out-of-Distribution Detection

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Yingsheng, Lu, Shuo, Liang, Jian, Zheng, Aihua, He, Ran
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915484438364160
author Wang, Yingsheng
Lu, Shuo
Liang, Jian
Zheng, Aihua
He, Ran
author_facet Wang, Yingsheng
Lu, Shuo
Liang, Jian
Zheng, Aihua
He, Ran
contents Out-of-distribution (OOD) detection helps models identify data outside the training categories, crucial for security applications. While feature-based post-hoc methods address this by evaluating data differences in the feature space without changing network parameters, they often require access to training data, which may not be suitable for some data privacy scenarios. This may not be suitable in scenarios where data privacy protection is a concern. In this paper, we propose a simple yet effective post-hoc method, termed Classifier-based Feature Reconstruction (ClaFR), from the perspective of subspace projection. It first performs an orthogonal decomposition of the classifier's weights to extract the class-known subspace, then maps the original data features into this subspace to obtain new data representations. Subsequently, the OOD score is determined by calculating the feature reconstruction error of the data within the subspace. Compared to existing OOD detection algorithms, our method does not require access to training data while achieving leading performance on multiple OOD benchmarks. Our code is released at https://github.com/Aie0923/ClaFR.
format Preprint
id arxiv_https___arxiv_org_abs_2509_06988
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Frustratingly Easy Feature Reconstruction for Out-of-Distribution Detection
Wang, Yingsheng
Lu, Shuo
Liang, Jian
Zheng, Aihua
He, Ran
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Out-of-distribution (OOD) detection helps models identify data outside the training categories, crucial for security applications. While feature-based post-hoc methods address this by evaluating data differences in the feature space without changing network parameters, they often require access to training data, which may not be suitable for some data privacy scenarios. This may not be suitable in scenarios where data privacy protection is a concern. In this paper, we propose a simple yet effective post-hoc method, termed Classifier-based Feature Reconstruction (ClaFR), from the perspective of subspace projection. It first performs an orthogonal decomposition of the classifier's weights to extract the class-known subspace, then maps the original data features into this subspace to obtain new data representations. Subsequently, the OOD score is determined by calculating the feature reconstruction error of the data within the subspace. Compared to existing OOD detection algorithms, our method does not require access to training data while achieving leading performance on multiple OOD benchmarks. Our code is released at https://github.com/Aie0923/ClaFR.
title Frustratingly Easy Feature Reconstruction for Out-of-Distribution Detection
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2509.06988