Research map / Label-efficient and robust learning / OOD and anomaly detection
Hierarchical Semi-Supervised Contrastive Learning for Contamination-Resistant Anomaly Detection
Gaoang Wang, Yibing Zhan, Xinchao Wang, Mingli Song, Klara Nahrstedt
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 sample-to-sample module (ablation) | HSCL · CIFAR-10 ablation | AUROC: 89.8 -> 97.1 | Improved |
| add sample-to-prototype module (ablation) | HSCL · CIFAR-10 ablation | AUROC: 82.9 -> 97.1 | Improved |
| add normal-to-abnormal module (ablation) | HSCL · CIFAR-10 ablation | AUROC: 94.1 -> 97.1 | Improved |
| add sample-to-sample module (ablation) | HSCL · CIFAR-100 ablation | AUROC: 72.9 -> 90.5 | Improved |
| add sample-to-prototype module (ablation) | HSCL · CIFAR-100 ablation | AUROC: 85.5 -> 90.5 | Improved |
| add normal-to-abnormal module (ablation) | HSCL · CIFAR-100 ablation | AUROC: 85.3 -> 90.5 | Improved |
| add sample-to-sample module (ablation) | HSCL · LSUN ablation | AUROC: 67.0 -> 88.6 | Improved |
| add sample-to-prototype module (ablation) | HSCL · LSUN ablation | AUROC: 70.3 -> 88.6 | Improved |
| add normal-to-abnormal module (ablation) | HSCL · LSUN ablation | AUROC: 87.1 -> 88.6 | Improved |
| add the positive term of the prototype loss (ablation) | HSCL · CIFAR-10 ablation | AUROC: 96.5 -> 97.1 | Improved |
| add the negative term of the prototype loss (ablation) | HSCL · CIFAR-10 ablation | AUROC: 96.1 -> 97.1 | Improved |
| increase number of prototypes K from 1 to 8 (ablation) | HSCL · CIFAR-10, CIFAR-100, LSUN | AUROC | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · better on 4 of 4 shared results
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
- Generating and Reweighting Dense Contrastive Patterns for Unsupervised Anomaly Detection (AAAI 2024 · similarity 0.84)
- Revisiting Graph Contrastive Learning on Anomaly Detection: A Structural Imbalance Perspective (AAAI 2025 · similarity 0.84)
- Contamination-Resilient Anomaly Detection via Adversarial Learning on Partially-Observed Normal and Anomalous Data (ICML 2024 · similarity 0.84)
- Normal-Abnormal Guided Generalist Anomaly Detection (NeurIPS 2025 · similarity 0.83)
- AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios (CVPR 2025 · similarity 0.83)
- CNC: Cross-modal Normality Constraint for Unsupervised Multi-class Anomaly Detection (AAAI 2025 · similarity 0.83)
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.
