Research map / Label-efficient and robust learning / OOD and anomaly detection
Single Layer Predictive Normalized Maximum Likelihood for Out-of-Distribution Detection
Koby Bibas, Meir Feder, Tal Hassner
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 pNML regret to Baseline | Baseline (DenseNet-BC-100) · CIFAR-100 vs Gaussian | AUROC: 30.6 -> 100 | Improved |
| add pNML regret to Baseline | Baseline (ResNet-34) · CIFAR-100 vs Gaussian | AUROC: 45.0 -> 86.5 | Improved |
| add pNML regret to ODIN | ODIN (DenseNet-BC-100) · CIFAR-100 vs LSUN (C) | AUROC: 91.5 -> 93.1 | Improved |
| add pNML regret to Gram | Gram (DenseNet-BC-100) · CIFAR-100 vs LSUN (C) | AUROC: 91.4 -> 94.5 | Improved |
| add pNML regret to OECC | OECC (DenseNet-BC-100) · CIFAR-100 vs LSUN (C) | AUROC: 93.9 -> 96.1 | Improved |
| add pNML regret to Gram | Gram (ResNet-34) · CIFAR-100 vs LSUN (C) | AUROC: 92.2 -> 94.6 | Improved |
| add pNML regret to OECC | OECC (ResNet-34) · CIFAR-100 vs LSUN (C) | AUROC: 95.7 -> 97.8 | Improved |
| add pNML regret to Energy | Energy (WideResNet-40) · CIFAR-100 vs ImageNet (R) | AUROC: 71.4 -> 87.0 | Improved |
| add pNML regret to Energy | Energy (WideResNet-40) · CIFAR-100 vs Gaussian | TNR at TPR 95%: 9.6 -> 100 | Improved |
| add pNML regret to Energy | Energy (WideResNet-40) · CIFAR-100 vs LSUN (C) | AUROC: 95.9 -> 95.5 | Hurt |
| add pNML regret to Energy | Energy (WideResNet-40) · CIFAR-10 vs ImageNet (R) | AUROC: 98.1 -> 98.1 | No clear effect |
| add pNML regret to Baseline | Baseline (ResNet-18) · ImageNet-30 vs CIFAR-10 | AUROC: 89.99 -> 99.79 | Improved |
Closest papers by meaning
- Optimal Parameter and Neuron Pruning for Out-of-Distribution Detection (NeurIPS 2023 · similarity 0.89)
- Deep Hybrid Models for Out-of-Distribution Detection (CVPR 2022 · similarity 0.88)
- A Statistical Framework for Efficient Out of Distribution Detection in Deep Neural Networks (ICLR 2022 · similarity 0.87)
- DOS: Diverse Outlier Sampling for Out-of-Distribution Detection (ICLR 2024 · similarity 0.87)
- Understanding Likelihood of Normalizing Flow and Image Complexity through the Lens of Out-of-Distribution Detection (AAAI 2024 · similarity 0.86)
- Leveraging Perturbation Robustness to Enhance Out-of-Distribution Detection (CVPR 2025 · similarity 0.86)
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.
