Research map / Label-efficient and robust learning / OOD and anomaly detection
WOOD: Wasserstein-Based Out-of-Distribution Detection
Yinan Wang, Wenbo Sun, Jionghua Judy Jin, Zhenyu Kong, Xiaowei Yue
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use WOOD (binary distance matrix) instead of ODIN | DenseNet-100 · CIFAR-10 vs SVHN | FNR at 95% TNR: 0.1591 -> 0.0005 | Improved |
| use WOOD (binary distance matrix) instead of Maha | DenseNet-100 · CIFAR-10 vs TinyImageNet-c | FNR at 95% TNR: 0.4878 -> <0.000001 | Improved |
| use WOOD (binary distance matrix) instead of ODIN | DenseNet-100 · MNIST vs FashionMNIST | AUROC: 0.741 -> >0.999999 | Improved |
| use WOOD (binary) instead of ODIN | DenseNet-100 · CIFAR-10 vs TinyImageNet-r | AUROC: 0.991 -> 0.988 | Mixed |
| switch WOOD from binary to dynamic distance matrix (ablation) | WOOD · CIFAR-10 vs SVHN | FNR at 95% TNR: 0.0005 -> 0.0046 | Hurt |
| switch WOOD from binary to dynamic distance matrix (ablation) | WOOD · CIFAR-10 vs TinyImageNet-r | FNR at 95% TNR: 0.0095 -> 0.0069 | Improved |
| use dynamic instead of binary distance matrix for score computation (ablation) | WOOD · CIFAR-10 score computation time per 100 samples | seconds: 0.0307 -> 0.0035 | Improved |
Closest papers by meaning
- CASE: Exploiting Intra-class Compactness and Inter-class Separability of Feature Embeddings for Out-of-Distribution Detection (AAAI 2024 · similarity 0.88)
- Exploiting Discrepancy in Feature Statistic for Out-of-Distribution Detection (AAAI 2024 · similarity 0.87)
- Learning to Shape In-distribution Feature Space for Out-of-distribution Detection (NeurIPS 2024 · similarity 0.87)
- Learning with Mixture of Prototypes for Out-of-Distribution Detection (ICLR 2024 · similarity 0.87)
- Enhancing the Power of OOD Detection via Sample-Aware Model Selection (CVPR 2024 · similarity 0.87)
- EAT: Towards Long-Tailed Out-of-Distribution Detection (AAAI 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
