Research map / Label-efficient and robust learning / OOD and anomaly detection
LoCoOp: Few-Shot Out-of-Distribution Detection via Prompt Learning
Atsuyuki Miyai, Qing Yu, Go Irie, Kiyoharu Aizawa
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 local-feature OOD regularization (LoCoOp) to CoOp, GL-MCM score, 16-shot | CoOp · ImageNet-1K OOD benchmarks | FPR95: 30.67 -> 28.66 | Improved |
| add local-feature OOD regularization, MCM score, 16-shot | CoOp · ImageNet-1K OOD benchmarks | FPR95: 36.83 -> 33.98 | Improved |
| add local-feature OOD regularization, 1-shot, GL-MCM | CoOp · ImageNet-1K OOD benchmarks | FPR95: 36.58 -> 33.52 | Improved |
| add OOD regularization on ResNet-50 backbone, 16-shot | CoOp_GL · ImageNet-1K OOD benchmarks | FPR95: 38.27 -> 34.45 | Improved |
| add OOD regularization on ImageNet-100 ID, 16-shot | CoOp_GL · ImageNet-100 OOD | FPR95: 13.82 -> 10.77 | Improved |
| use LoCoOp instead of CoCoOp | prompt learning OOD detection · ImageNet-1K OOD, 16-shot | FPR95 and inference time: 35.53 -> 33.98 FPR95; 149 ms -> 2.59 ms | Improved |
| add OOD regularization | CoOp · ImageNet-1K val, 16-shot | top-1 ID accuracy: 72.1 -> 71.7 | Hurt |
| use rank-based top-K extraction of ID-irrelevant regions instead of entropy-based thresholding (ablation) | LoCoOp_GL · ImageNet-1K OOD, 16-shot | FPR95: 36.38 -> 28.66 | Improved |
| use rank-based instead of probability-based region extraction (ablation) | LoCoOp_GL · ImageNet-1K OOD, 16-shot | FPR95: 28.65 -> 28.66 | No clear effect |
| treat all local regions as OOD (K=0) instead of K=200 (ablation) | LoCoOp · ImageNet-1K OOD, 16-shot | detection performance | Hurt |
| vary K between 200 and 980 (ablation) | LoCoOp · ImageNet-1K OOD, 16-shot | detection performance | No clear effect |
| set lambda below 1 (0.25) instead of 1 (ablation) | LoCoOp · ImageNet-1K OOD, 16-shot | FPR95 / AUROC | Improved |
Later papers that compared with this method
11 later accepted papers report LoCoOp as a baseline in their results tables; 11 of them report a better number on most of the results they share with it.
- Intra-Image Mining and Symmetric Maximum Concept Matching for Few Shot Out-of-Distribution Detection (AAAI 2026 · better on 4 of 5 shared results)
- A Close Look at Negative Label Guided Out-of-distribution Detection in Pre-trained Vision-Language Models (ICML 2026 · better on 2 of 2 shared results)
- STAR: Test-Time Adaptation Can Enhance Universal Prompt Learning for Vision-Language Models (CVPR 2026 · better on 2 of 2 shared results)
- A Novel Fine-Tuned CLIP-OOD Detection Method with Double Loss Constraint Through Optimal Transport Semantic Alignment (AAAI 2026 · better on 2 of 2 shared results)
- Local-Prompt: Extensible Local Prompts for Few-Shot Out-of-Distribution Detection (ICLR 2025 · better on 7 of 7 shared results)
- Overcoming Shortcut Problem in VLM for Robust Out-of-Distribution Detection (CVPR 2025 · better on 2 of 2 shared results)
- Auxiliary Prompt Tuning of Vision-Language Models for Few-Shot Out-of-Distribution Detection (ICCV 2025 · better on 2 of 2 shared results)
- Adaptive Multi-prompt Contrastive Network for Few-shot Out-of-distribution Detection (ICML 2025 · better on 1 of 1 shared result)
3 more not listed.
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Adaptive Multi-prompt Contrastive Network for Few-shot Out-of-distribution Detection (ICML 2025 · similarity 0.90)
- Local-Prompt: Extensible Local Prompts for Few-Shot Out-of-Distribution Detection (ICLR 2025 · similarity 0.89)
- Conjugated Semantic Pool Improves OOD Detection with Pre-trained Vision-Language Models (NeurIPS 2024 · similarity 0.89)
- ID-like Prompt Learning for Few-Shot Out-of-Distribution Detection (CVPR 2024 · similarity 0.88)
- Revisiting Few-Shot Object Detection with Vision-Language Models (NeurIPS 2024 · similarity 0.87)
- Self-Calibrated Tuning of Vision-Language Models for Out-of-Distribution Detection (NeurIPS 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.
