Research map / Label-efficient and robust learning / OOD and anomaly detection
Geometrically Constrained Outlier Synthesis
Daniil Karzanov, Marcin Detyniecki
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 GCOS conformal geometric synthesis instead of VOS Gaussian synthesis | energy-based OOD detector, WRN-40-2 · MVTec near-OOD | FPR95: 70.77 -> 23.08 | Improved |
| use GCOS instead of VOS | energy-based OOD detector · Colored MNIST | FPR95: 18.50 -> 1.00 | Improved |
| use GCOS instead of VOS | energy-based OOD detector · Stanford Dogs | AUROC: 99.25 -> 99.55 | Improved |
| use GCOS instead of VOS | energy-based OOD detector · Retinopathy | FPR95: 80.00 -> 73.00 | Improved |
| add VOS Gaussian outlier synthesis to no-regularization baseline | energy-based OOD detector · MVTec near-OOD | FPR95: 27.69 -> 70.77 | Hurt |
| add VOS Gaussian outlier synthesis to no-regularization baseline | energy-based OOD detector · Colored MNIST | FPR95: 32.50 -> 18.50 | Improved |
| use GCOS instead of NCIS (diffusion-based) | energy-based OOD detector · MVTec near-OOD | AUROC: 96.50 -> 95.61 | Hurt |
| use GCOS instead of NCIS | energy-based OOD detector · Average over 4 datasets | Avg. AUROC: 91.97 -> 93.47 | Improved |
| replace energy-based inference with conformal hypothesis testing head (ablation) | GCOS + L_uncertainty · MVTec near-OOD | AUROC: 94.86 -> 50.92 | Hurt |
| replace energy-based inference with conformal hypothesis testing head (ablation) | GCOS + L_reg(Mahalanobis) · Colored MNIST | FPR95: 1.50 | Improved |
| use energy regularization with conformal testing head (ablation) | GCOS + L_reg(Energy) · MVTec near-OOD | AUROC: 63.67 -> 70.62 | Improved |
Closest papers by meaning
- Out-of-Distribution Detection with Prototypical Outlier Proxy (AAAI 2025 · similarity 0.88)
- A Systematic Analysis of Out-of-Distribution Detection Under Representation and Training Paradigm Shifts (NeurIPS 2026 · similarity 0.87)
- Extremely Simple Multimodal Outlier Synthesis for Out-of-Distribution Detection and Segmentation (NeurIPS 2025 · similarity 0.87)
- BOOD: Boundary-based Out-Of-Distribution Data Generation (ICML 2025 · similarity 0.87)
- RODEO: Robust Outlier Detection via Exposing Adaptive Out-of-Distribution Samples (ICML 2024 · similarity 0.87)
- MahaVar: OOD Detection via Class-wise Mahalanobis Distance Variance under Neural Collapse (NeurIPS 2026 · 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.
