Research map / Label-efficient and robust learning / OOD and anomaly detection
Uncertainty-Aware Optimal Transport for Semantically Coherent Out-of-Distribution Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 uncertainty-aware ET with inter-cluster extension | UDG on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 36.22 -> 8.53 | Helped |
| Use uncertainty-aware ET with inter-cluster extension | UDG on the CIFAR-100 Benchmark · CIFAR-100 Benchmark | FPR95: 75.45 -> 41.05 | Helped |
| Add L_rep to UDG (ablation) | UDG on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 21.57 -> 17.75 | Helped |
| Replace UDG's IDF strategy with ET (ablation) | UDG on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 21.57 -> 11.39 | Helped |
| Add L_rep to ET (ablation) | ET on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 11.39 -> 8.53 | Helped |
| Use energy uncertainty rather than variance to guide transport (ablation) | ET on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 13.46 -> 8.53 | Helped |
| Use energy uncertainty rather than entropy to guide transport (ablation) | ET on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 10.94 -> 8.53 | Helped |
| Use T-energy instead of MSP as the OOD score (ablation) | Ours on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 13.55 -> 8.53 | Helped |
| Use T-energy instead of energy as the OOD score (ablation) | Ours on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 12.86 -> 8.53 | Helped |
| Perform ET throughout training rather than only during the last 50 epochs (ablation) | Ours on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 13.54 -> 8.53 | Helped |
| Perform ET throughout training rather than only during the last 150 epochs (ablation) | Ours on the CIFAR-10 Benchmark · CIFAR-10 Benchmark | FPR95: 11.83 -> 8.53 | Helped |
Closest papers by meaning
- The Best of Both Worlds: On the Dilemma of Out-of-distribution Detection (NeurIPS 2024 · similarity 0.88)
- ImageNet-OOD: Deciphering Modern Out-of-Distribution Detection Algorithms (ICLR 2024 · similarity 0.87)
- Diffusion-based Layer-wise Semantic Reconstruction for Unsupervised Out-of-Distribution Detection (NeurIPS 2024 · similarity 0.87)
- RUNA: Object-Level Out-of-Distribution Detection via Regional Uncertainty Alignment of Multimodal Representations (AAAI 2025 · similarity 0.86)
- A Systematic Analysis of Out-of-Distribution Detection Under Representation and Training Paradigm Shifts (NeurIPS 2026 · similarity 0.86)
- Rethinking the Evaluation of Out-of-Distribution Detection: A Sorites Paradox (NeurIPS 2024 · 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.