Research map / Detection and segmentation / Open-set and few-shot detection
Object-Centric Auto-Encoders and Dummy Anomalies for Abnormal Event Detection in Video
Check 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 |
|---|---|---|---|
| Add the k-means and one-versus-rest SVM classification approach to frame-level CAE features (ablation) | Abnormal event detection · ShanghaiTech | frame-level AUC: 72.4 -> 78.7 | Helped |
| Use object-centric CAE features instead of frame-level CAE features (ablation) | One-versus-rest SVM abnormal event detection · ShanghaiTech | frame-level AUC: 78.7 -> 84.9 | Helped |
| Learn object-centric CAE features instead of using pre-trained SSD features (ablation) | One-versus-rest SVM abnormal event detection · ShanghaiTech | frame-level AUC: 81.3 -> 84.9 | Helped |
| Add object-centric motion CAE features (ablation) | Appearance CAE features + one-versus-rest SVM · ShanghaiTech | frame-level AUC: 82.2 -> 84.9 | Helped |
| Add object-centric appearance CAE features (ablation) | Motion CAE features + one-versus-rest SVM · ShanghaiTech | frame-level AUC: 83.0 -> 84.9 | Helped |
| Use k-means and one-versus-rest SVM instead of one-class SVM (ablation) | Combined object-centric CAE features · ShanghaiTech | frame-level AUC: 79.2 -> 84.9 | Helped |
| Vary the number of k-means clusters across 5, 10, 15, 20, 25, and 30 (ablation) | The proposed method · ShanghaiTech | frame-level AUC: variations are lower than 1.1% | No clear effect |
| Vary the SVM regularization parameter across 0.1, 1, 10, and 100 (ablation) | The proposed method · ShanghaiTech | frame-level AUC: performance variation is lower than 0.3% | No clear effect |
Closest papers by meaning
- Self-Distilled Masked Auto-Encoders are Efficient Video Anomaly Detectors (CVPR 2024 · similarity 0.90)
- Anomaly Detection in Video via Self-Supervised and Multi-Task Learning (CVPR 2021 · similarity 0.89)
- Learning Memory-Guided Normality for Anomaly Detection (CVPR 2020 · similarity 0.88)
- Prompt-Enhanced Multiple Instance Learning for Weakly Supervised Video Anomaly Detection (CVPR 2024 · similarity 0.87)
- Track Any Anomalous Object:A Granular Video Anomaly Detection Pipeline (CVPR 2025 · similarity 0.86)
- Learning Event Completeness for Weakly Supervised Video Anomaly Detection (ICML 2025 · similarity 0.85)
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.