Out-Of-Distribution Detection with Diversification (Provably)
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909398272573440 |
|---|---|
| author | Yao, Haiyun Han, Zongbo Fu, Huazhu Peng, Xi Hu, Qinghua Zhang, Changqing |
| author_facet | Yao, Haiyun Han, Zongbo Fu, Huazhu Peng, Xi Hu, Qinghua Zhang, Changqing |
| contents | Out-of-distribution (OOD) detection is crucial for ensuring reliable deployment of machine learning models. Recent advancements focus on utilizing easily accessible auxiliary outliers (e.g., data from the web or other datasets) in training. However, we experimentally reveal that these methods still struggle to generalize their detection capabilities to unknown OOD data, due to the limited diversity of the auxiliary outliers collected. Therefore, we thoroughly examine this problem from the generalization perspective and demonstrate that a more diverse set of auxiliary outliers is essential for enhancing the detection capabilities. However, in practice, it is difficult and costly to collect sufficiently diverse auxiliary outlier data. Therefore, we propose a simple yet practical approach with a theoretical guarantee, termed Diversity-induced Mixup for OOD detection (diverseMix), which enhances the diversity of auxiliary outlier set for training in an efficient way. Extensive experiments show that diverseMix achieves superior performance on commonly used and recent challenging large-scale benchmarks, which further confirm the importance of the diversity of auxiliary outliers. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_14049 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Out-Of-Distribution Detection with Diversification (Provably) Yao, Haiyun Han, Zongbo Fu, Huazhu Peng, Xi Hu, Qinghua Zhang, Changqing Machine Learning Computer Vision and Pattern Recognition Out-of-distribution (OOD) detection is crucial for ensuring reliable deployment of machine learning models. Recent advancements focus on utilizing easily accessible auxiliary outliers (e.g., data from the web or other datasets) in training. However, we experimentally reveal that these methods still struggle to generalize their detection capabilities to unknown OOD data, due to the limited diversity of the auxiliary outliers collected. Therefore, we thoroughly examine this problem from the generalization perspective and demonstrate that a more diverse set of auxiliary outliers is essential for enhancing the detection capabilities. However, in practice, it is difficult and costly to collect sufficiently diverse auxiliary outlier data. Therefore, we propose a simple yet practical approach with a theoretical guarantee, termed Diversity-induced Mixup for OOD detection (diverseMix), which enhances the diversity of auxiliary outlier set for training in an efficient way. Extensive experiments show that diverseMix achieves superior performance on commonly used and recent challenging large-scale benchmarks, which further confirm the importance of the diversity of auxiliary outliers. |
| title | Out-Of-Distribution Detection with Diversification (Provably) |
| topic | Machine Learning Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2411.14049 |