Research map / Video / Video segmentation and tracking
Learning Memory-Guided Normality for Anomaly Detection
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 memory module (ablation) | Ours-P · UCSD Ped2 | AUC (%): 94.3 -> 97.0; 2.7% | Helped |
| add the memory module (ablation) | Ours-P · CUHK Avenue | AUC (%): 84.5 -> 88.5; 4.0% | Helped |
| add the memory module (ablation) | Ours-P · ShanghaiTech | AUC (%): 66.8 -> 70.5; 3.7% | Helped |
| add the feature separateness loss (ablation) | Ours-R with memory and compactness loss · UCSD Ped2 | AUC (%): 86.4 -> 90.2 | Helped |
| add the feature separateness loss (ablation) | Ours-P with memory and compactness loss · UCSD Ped2 | AUC (%): 94.8 -> 97.0 | Helped |
| add the feature compactness loss (ablation) | Ours-R with memory and separateness loss · UCSD Ped2 | AUC (%): 89.3 -> 90.2 | Helped |
| add the feature compactness loss (ablation) | Ours-P with memory and separateness loss · UCSD Ped2 | AUC (%): 96.5 -> 97.0 | Helped |
| add feature compactness loss without feature separateness loss (ablation) | Ours-R with memory · UCSD Ped2 | AUC (%): 86.9 -> 86.4 | Hurt |
| add feature compactness loss without feature separateness loss (ablation) | Ours-P with memory · UCSD Ped2 | AUC (%): 95.0 -> 94.8 | Hurt |
| enable regular-score-gated memory updates at test time (ablation) | Ours-R · UCSD Ped2 | AUC (%): 87.1 -> 90.2 | Helped |
| enable regular-score-gated memory updates at test time (ablation) | Ours-P · UCSD Ped2 | AUC (%): 96.0 -> 97.0 | Helped |
| combine PSNR with memory-feature distance in the abnormality score (ablation) | Ours-R · UCSD Ped2 | AUC (%): 89.0 -> 90.2 | Helped |
Closest papers by meaning
- Dual Memory Units with Uncertainty Regulation for Weakly Supervised Video Anomaly Detection (AAAI 2023 · similarity 0.90)
- Holmes-VAU: Towards Long-term Video Anomaly Understanding at Any Granularity (CVPR 2025 · similarity 0.88)
- Prompt-Enhanced Multiple Instance Learning for Weakly Supervised Video Anomaly Detection (CVPR 2024 · similarity 0.87)
- Open-Vocabulary Video Anomaly Detection (CVPR 2024 · similarity 0.87)
- Harnessing Large Language Models for Training-free Video Anomaly Detection (CVPR 2024 · similarity 0.87)
- MULDE: Multiscale Log-Density Estimation via Denoising Score Matching for Video Anomaly Detection (CVPR 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.