Research map / Label-efficient and robust learning / OOD and anomaly detection
A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework
Weixin Luo, Wen Liu, Shenghua Gao
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 |
|---|---|---|---|
| learn all parameters in a stacked RNN instead of fixed-hyperparameter sparse coding (TSC) | TSC · UCSD Ped2 | frame-level AUC: 91.03% -> 92.21% | Improved |
| learn all parameters in a stacked RNN instead of TSC | TSC · CUHK Avenue | frame-level AUC: 80.56% -> 81.71% | Improved |
| learn all parameters in a stacked RNN instead of TSC | TSC · ShanghaiTech Campus | frame-level AUC: 67.94% -> 68.00% | No clear effect |
| use sRNN instead of Conv-AE | sRNN · UCSD Ped2 | frame-level AUC: 81.1% -> 92.21% | Improved |
| use sRNN instead of Conv-AE | sRNN · Moving-MNIST (synthetic) | frame-level AUC: 74.3% -> 86.52% | Improved |
| use TSC instead of Conv-AE | TSC · ShanghaiTech Campus | frame-level AUC: 60.85% -> 67.94% | Improved |
| use sRNN instead of ConvLSTM auto-encoder | sRNN · UCSD Ped2 | frame-level AUC: 88.10% -> 92.21% | Improved |
| use sRNN instead of LSTM auto-encoder | sRNN · CUHK Avenue | frame-level AUC: 75.33% -> 81.71% | Improved |
| increase stacked RNN layers from 3 to 30 (more sparsity) (ablation) | sRNN · UCSD Ped2 | frame-level AUC: 91.03% -> 88.10% | Hurt |
| increase sparsity from 80.0% to 92.7% via more layers (ablation) | sRNN · UCSD Ped2 | sparsity: 80.0% -> 92.7% | Mixed |
| use a small temporal coherence weight lambda2 (ablation) | TSC · Avenue and Ped2 | frame-level AUC | Hurt |
| increase sparsity weight lambda1 (ablation) | TSC · Avenue and Ped2 | frame-level AUC | No clear effect |
Closest papers by meaning
- Video Anomaly Detection with Sparse Coding Inspired Deep Neural Networks (TPAMI 2019 · similarity 0.95)
- Time Series Anomaly Detection with Multiresolution Ensemble Decoding (AAAI 2021 · similarity 0.85)
- Noise-Resistant Video Anomaly Detection via RGB Error-Guided Multiscale Predictive Coding and Dynamic Memory (CVPR 2025 · similarity 0.83)
- A Diffusion-Based Framework for Multi-Class Anomaly Detection (AAAI 2024 · similarity 0.83)
- UCF-Crime-DVS: A Novel Event-Based Dataset for Video Anomaly Detection with Spiking Neural Networks (AAAI 2025 · similarity 0.82)
- Sparse Spiking Neural Network: Exploiting Heterogeneity in Timescales for Pruning Recurrent SNN (ICLR 2024 · similarity 0.82)
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.
