Research map / Video / Video segmentation and tracking
Self-Trained Deep Ordinal Regression for End-to-End Video Anomaly Detection
Guansong Pang, Cheng Yan, Chunhua Shen, Anton van den Hengel, Xiao Bai
Check 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 |
|---|---|---|---|
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UCSD Ped1 | frame-level AUC: 56.3% -> 71.7% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UCSD Ped2 | frame-level AUC: 67.5% -> 83.2% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · Subway Entrance | frame-level AUC: 80.5% -> 88.1% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · Subway Exit | frame-level AUC: 91.0% -> 92.7% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UMN Scene1 | frame-level AUC: 87.3% -> 99.9% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UMN Scene2 | frame-level AUC: 88.1% -> 99.9% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UMN Scene3 | frame-level AUC: 91.5% -> 99.7% | Improved |
| Replace the initial Sp + iForest detector with the self-trained ordinal-regression method | Video anomaly detection · UMN All Scenes | frame-level AUC: 87.1% -> 97.4% | Improved |
| Replace ResNet-50 with VGG (ablation) | Self-trained deep ordinal regression · UCSD Ped1 | AUC: 71.7% -> 70.4% | Hurt |
| Replace ResNet-50 with 3DConv (ablation) | Self-trained deep ordinal regression · UCSD Ped2 | AUC: 83.2% -> 82.6% | Hurt |
| Replace ResNet-50 with VGG (ablation) | Self-trained deep ordinal regression · Subway Entrance | AUC: 88.1% -> 86.5% | Hurt |
| Replace ResNet-50 with 3DConv (ablation) | Self-trained deep ordinal regression · Subway Exit | AUC: 92.7% -> 93.6% | Improved |
Closest papers by meaning
- Real-World Anomaly Detection in Surveillance Videos (CVPR 2018 · similarity 0.90)
- Anomaly Detection in Video via Self-Supervised and Multi-Task Learning (CVPR 2021 · similarity 0.89)
- A Discriminative Framework for Anomaly Detection in Large Videos (ECCV 2016 · similarity 0.89)
- EVAL: Explainable Video Anomaly Localization (CVPR 2023 · similarity 0.88)
- UBnormal: New Benchmark for Supervised Open-Set Video Anomaly Detection (CVPR 2022 · similarity 0.88)
- PANDA: Towards Generalist Video Anomaly Detection via Agentic AI Engineer (NeurIPS 2025 · 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.
