Research map / Label-efficient and robust learning / OOD and anomaly detection
Let OOD Feature Exploring Vast Predefined Classifiers
Kewen Xia, Xiaodong Yue, WeiZhipeng, Yaxin Peng, Zihao Li, Jianxiang Zhu, Jie Shi, PeilinXu
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 |
|---|---|---|---|
| add L_VEBV loss to L_ENC + L_OC (ablation) | VPC · CIFAR-10 | FPR95: 2.68 -> 2.27 | Improved |
| add L_VEBV loss to L_ENC + L_OC (ablation) | VPC · CIFAR-100 | FPR95: 34.89 -> 32.04 | Improved |
| use evidential neural collapse + orthogonality instead of CE + OE (ablation) | VPC · CIFAR-10 | FPR95: 3.44 -> 2.68 | Improved |
| use evidential neural collapse + orthogonality instead of CE + OE (ablation) | VPC · CIFAR-100 | FPR95: 36.14 -> 34.89 | Improved |
| replace L_NC with L_ENC (ablation) | VPC · CIFAR-100 | FPR95: 34.65 -> 34.89 | No clear effect |
| increase VEBV subspace size from 10 to 2000 (ablation) | VPC · CIFAR-10 | FPR95: 2.68 -> 2.21 | Improved |
| increase VEBV subspace size from 100 to 1000 (ablation) | VPC · CIFAR-100 | FPR95: 34.89 -> 32.04 | Improved |
| increase VEBV subspace size from 1000 to 2000 (ablation) | VPC · CIFAR-100 | FPR95: 32.04 -> 32.56 | Hurt |
| use VPC Score instead of MSP (ablation) | VPC · CIFAR-100 (ResNet-18) | FPR95: 44.75 -> 38.96 | Improved |
| use VPC Score instead of MSP (ablation) | VPC · CIFAR-10 (ResNet-18) | AUROC: 98.81 -> 98.32 | Hurt |
| use VPC instead of PFS | ResNet-50 on ImageNet-1k · ImageNet-1k | average FPR95: 57.85 -> 56.50 | Improved |
| use VPC instead of PFS | ViT-B-16 on ImageNet-1k · ImageNet-1k | average FPR95: 56.93 -> 56.20 | Improved |
Closest papers by meaning
- Is Complex Training Necessary for Long-Tailed OOD Detection? A Re-think from Feature Geometry (NeurIPS 2026 · similarity 0.88)
- Learning to Shape In-distribution Feature Space for Out-of-distribution Detection (NeurIPS 2024 · similarity 0.88)
- Position: Supervised Classifiers Answer the Wrong Questions for OOD Detection (ICML 2025 · similarity 0.88)
- EAT: Towards Long-Tailed Out-of-Distribution Detection (AAAI 2024 · similarity 0.88)
- Long-Tailed Out-of-Distribution Detection: Prioritizing Attention to Tail (AAAI 2025 · similarity 0.88)
- Mining In-distribution Attributes in Outliers for Out-of-distribution Detection (AAAI 2025 · similarity 0.88)
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.
