Research map / Label-efficient and robust learning / OOD and anomaly detection
Semantically Coherent Out-of-Distribution Detection
Jingkang Yang, Haoqi Wang, Litong Feng, Xiaopeng Yan, Huabin Zheng, Wayne Zhang, Ziwei Liu
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 IDF operator F (ablation) | OE with L_A · SC-OOD CIFAR-10 ablation | FPR95: 51.41 -> 36.22 | Improved |
| add IDF operator F (ablation) | OE (L_C^I + L_C^U) · SC-OOD CIFAR-10 ablation | FPR95: 50.53 -> 40.93 | Improved |
| add IDF operator F (ablation) | OE (L_C^I + L_C^U) · SC-OOD CIFAR-10 ablation | ACC: 91.87 -> 92.34 | Improved |
| add auxiliary clustering loss L_A after IDF (ablation) | L_C^I + L_C^U + F · SC-OOD CIFAR-10 ablation | FPR95: 40.93 -> 36.22 | Improved |
| add auxiliary clustering loss L_A to classification on CIFAR only (ablation) | L_C^I · SC-OOD CIFAR-10 ablation | FPR95: 58.27 -> 55.62 | Improved |
| add L_A to classification with OOD-mixed unlabeled data (ablation) | L_C^I · SC-OOD CIFAR-10 ablation | FPR95: 55.62 -> 62.75 | Hurt |
| add L_A to OE without IDF (ablation) | OE (L_C^I + L_C^U) · SC-OOD CIFAR-10 ablation | FPR95: 50.53 -> 51.41 | No clear effect |
| add unlabeled-data OE loss L_C^U (ablation) | L_C^I · SC-OOD CIFAR-10 ablation | FPR95: 58.27 -> 50.53 | Improved |
| add unlabeled-data OE loss L_C^U (ablation) | L_C^I · SC-OOD CIFAR-10 ablation | ACC: 94.94 -> 91.87 | Hurt |
| add unlabeled Tiny-ImageNet to pure deep clustering (ablation) | L_A · SC-OOD CIFAR-10 ablation | FPR95: 89.53 -> 91.15 | Hurt |
| use group-based IDF (UDG) instead of sample-based SORT/THRESH (ablation) | IDF strategy · SC-OOD CIFAR-10 | FPR95 | Improved |
| increase number of clusters K to >=500 (ablation) | UDG · SC-OOD CIFAR-10 | FPR95 | No clear effect |
Closest papers by meaning
- ImageNet-OOD: Deciphering Modern Out-of-Distribution Detection Algorithms (ICLR 2024 · similarity 0.93)
- Rethinking the Evaluation of Out-of-Distribution Detection: A Sorites Paradox (NeurIPS 2024 · similarity 0.92)
- Image Background Serves as Good Proxy for Out-of-distribution Data (ICLR 2024 · similarity 0.92)
- A Systematic Analysis of Out-of-Distribution Detection Under Representation and Training Paradigm Shifts (NeurIPS 2026 · similarity 0.91)
- EAT: Towards Long-Tailed Out-of-Distribution Detection (AAAI 2024 · similarity 0.90)
- Is Complex Training Necessary for Long-Tailed OOD Detection? A Re-think from Feature Geometry (NeurIPS 2026 · similarity 0.90)
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.
