Research map / Label-efficient and robust learning / OOD and anomaly detection
Detection of unknown unknowns in autonomous systems
Ayan Banerjee, Sandeep Gupta
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 LTC-NN to SINDY-MPC (ablation) | SPIE-AD · F8Stuck (synthetic U2, no PA) | F1: 90.4 -> 92 | Improved |
| add LTC-NN to SINDY-MPC (ablation) | SPIE-AD · MSL / real-world datasets | F1: 74 -> 86 | Improved |
| use point adjustment (PA) at evaluation (ablation) | Anomaly Transformer · SMD/SMAP/MSL, data leakage allowed | F1: 25 -> 90 | Improved |
| remove validation set containing anomalies (zero-shot) (ablation) | Anomaly Transformer · SMD/SMAP/MSL | F1: 25 -> 0 | Hurt |
| remove validation set containing anomalies (zero-shot) (ablation) | SPIE-AD + LTCNN · SMD/SMAP/MSL | F1: 84 -> 82 | No clear effect |
| replace LTC-NN with NODE (ablation) | SPIE-AD · real-world U2 data | execution time: 1.2x faster | Mixed |
| replace LTC-NN with CT-RNN (ablation) | SPIE-AD · real-world U2 data | execution time: 1.7x faster | Mixed |
| use SINDY-MPC instead of W-SINDY or SINDY without control (ablation) | SPIE-AD · U2 benchmarks | all metrics | Improved |
| add a sine term to the SINDY library (ablation) | SPIE-AD · U2 benchmarks | all metrics | Hurt |
| increase miscoverage level alpha (ablation) | SPIE-ADL · F8 example | all metrics: coverage 0.95 to 0.2 | Hurt |
| reduce sampling frequency from 120 Hz to 20 Hz (ablation) | SPIE-ADL · U2 benchmarks | all metrics: 20 Hz to 120 Hz sweep | Hurt |
| raise noise from 20 dB to 5 dB SNR (ablation) | SPIE-AD · U2 benchmarks | accuracy, precision, F1: 20 dB -> 5 dB | Hurt |
Closest papers by meaning
- The Elephant in the Room: Towards A Reliable Time-Series Anomaly Detection Benchmark (NeurIPS 2024 · similarity 0.87)
- Supervision Recovery for Time Series Anomaly Detection via Counterfactual Pairing (NeurIPS 2026 · similarity 0.86)
- Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection (NeurIPS 2025 · similarity 0.84)
- Detect, Explain, Interpret: An End-to-End Benchmark for Time Series Anomaly Detection, Explainability and Interpretability. (NeurIPS 2026 · similarity 0.84)
- When Model Meets New Normals: Test-Time Adaptation for Unsupervised Time-Series Anomaly Detection (AAAI 2024 · similarity 0.84)
- KAN-AD: Time Series Anomaly Detection with Kolmogorov–Arnold Networks (ICML 2025 · similarity 0.84)
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.
