Hyperbolic Metric Learning for Visual Outlier Detection
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_ | 1866910618918846464 |
|---|---|
| author | Gonzalez-Jimenez, Alvaro Lionetti, Simone Bazazian, Dena Gottfrois, Philippe Gröger, Fabian Pouly, Marc Navarini, Alexander |
| author_facet | Gonzalez-Jimenez, Alvaro Lionetti, Simone Bazazian, Dena Gottfrois, Philippe Gröger, Fabian Pouly, Marc Navarini, Alexander |
| contents | Out-Of-Distribution (OOD) detection is critical to deploy deep learning models in safety-critical applications. However, the inherent hierarchical concept structure of visual data, which is instrumental to OOD detection, is often poorly captured by conventional methods based on Euclidean geometry. This work proposes a metric framework that leverages the strengths of Hyperbolic geometry for OOD detection. Inspired by previous works that refine the decision boundary for OOD data with synthetic outliers, we extend this method to Hyperbolic space. Interestingly, we find that synthetic outliers do not benefit OOD detection in Hyperbolic space as they do in Euclidean space. Furthermore we explore the relationship between OOD detection performance and Hyperbolic embedding dimension, addressing practical concerns in resource-constrained environments. Extensive experiments show that our framework improves the FPR95 for OOD detection from 22\% to 15\% and from 49% to 28% on CIFAR-10 and CIFAR-100 respectively compared to Euclidean methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_15260 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Hyperbolic Metric Learning for Visual Outlier Detection Gonzalez-Jimenez, Alvaro Lionetti, Simone Bazazian, Dena Gottfrois, Philippe Gröger, Fabian Pouly, Marc Navarini, Alexander Computer Vision and Pattern Recognition Out-Of-Distribution (OOD) detection is critical to deploy deep learning models in safety-critical applications. However, the inherent hierarchical concept structure of visual data, which is instrumental to OOD detection, is often poorly captured by conventional methods based on Euclidean geometry. This work proposes a metric framework that leverages the strengths of Hyperbolic geometry for OOD detection. Inspired by previous works that refine the decision boundary for OOD data with synthetic outliers, we extend this method to Hyperbolic space. Interestingly, we find that synthetic outliers do not benefit OOD detection in Hyperbolic space as they do in Euclidean space. Furthermore we explore the relationship between OOD detection performance and Hyperbolic embedding dimension, addressing practical concerns in resource-constrained environments. Extensive experiments show that our framework improves the FPR95 for OOD detection from 22\% to 15\% and from 49% to 28% on CIFAR-10 and CIFAR-100 respectively compared to Euclidean methods. |
| title | Hyperbolic Metric Learning for Visual Outlier Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.15260 |