Rank-1 权重校正的可训练权重平均方法用于 OOD 检测
Rank-1 Weight-Corrected Trainable Weight Averaging for Out-of-Distribution Detection
AI 导读
研究者在可训练权重平均（TWA）基础上，用 SVD 提取的 rank-1 主导权重分量作为校正方向，其贡献通过分布内分类目标学习，以改进 OOD 检测。
正文 · 原文
Abstract:Out-of-distribution (OOD) detection identifies inputs outside the training distribution. Rank-based analyses such as RankFeat and RankWeight suggest that OOD overconfidence is closely related to dominant low-rank components in feature and weight spaces. Since weight averaging operates directly in parameter space and improves model generalization, we consider it for OOD detection from a dominant-component perspective. Trainable Weight Averaging (TWA) assigns a single coefficient to each checkpoint, limiting independent control of dominant weight components. We extend TWA with leading rank-one weight components extracted by Singular Value Decomposition as correction directions, whose contributions are learned through the in-distribution classification objective. On ResNet50, our method improves over fine-tuning and standard TWA across the evaluated datasets with MSP and RankWeight, both with and without Outlier Exposure. Mahalanobis-based evaluation yields mixed results, indicating that the benefit depends on the OOD scoring protocol. Target-layer analyses with MSP and Energy favor correction near the classifier.
| Comments: | Accepted to ACCV2026 Workshop on Subspaces in the Era of Foundation Models |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.33909 [cs.CV] |
| (or arXiv:2610.33909v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.33909 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hiroaki Aizawa [view email]
[v1]
Sat, 24 Oct 2026 20:45:04 UTC (336 KB)
来源：arXiv cs.AI · arxiv.org