Research map / Time series / Anomaly detection
ICDiffAD: Implicit Conditioning Diffusion Model for Time Series Anomaly Detection
Fan Zhang, Sinchee Chin, Jing-Hao Xue, Wenming Yang
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 implicit conditioning (ablation) | vanilla diffusion TSAD · Average over 5 datasets | F1 (%): 36.52 -> 40.27 | Improved |
| add SNR scheduler on top of IC (ablation) | ICDiffAD · Average over 5 datasets | F1 (%): 40.27 -> 42.74 | Improved |
| add SIC corruption/step estimation (ablation) | ICDiffAD · Average over 5 datasets | F1 (%): 42.74 -> 43.81 | Improved |
| add SIC (full model) on PSM (ablation) | ICDiffAD · PSM | F1 (%): 54.80 -> 56.87 | Improved |
| replace linear g(t) with quadratic (ablation) | SNR scheduler · MSL | F1 (%): 33.17 -> 34.78 | No clear effect |
| replace linear g(t) with cosine (ablation) | SNR scheduler · SMAP | F1 (%): 28.01 -> 26.76 | Hurt |
| replace linear g(t) with quadratic (ablation) | SNR scheduler · SMD | F1 (%): 23.68 -> 27.42 | Improved |
| lower target SNR (TSNRdB) during training (ablation) | SNR scheduler · MSL | F1: 0.2564 -> 0.3589 | Improved |
| lower target SNR (TSNRdB) during training (ablation) | SNR scheduler · SMD | F1: 0.1725 -> 0.2437 | Improved |
| increase Gaussian filter sigma (ablation) | SIC module · SMAP | precision / recall: P = 0.21 -> 0.67; R = 0.89 -> 0.34 | Mixed |
| increase Gaussian filter sigma (ablation) | SIC module · SMD | precision / recall: P = 0.18 -> 0.52; R = 0.91 -> 0.41 | Mixed |
| increase sigma beyond 1 (ablation) | SIC module · PSM | performance: P = 0.45, R = 0.76 at sigma = 1 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- SARAD — SARAD: Spatial Association-Aware Anomaly Detection and Diagnosis for Multivariate Time Series (NeurIPS 2024) · better on 4 of 6 shared results
- TimesNet — TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis (ICLR 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Supervision Recovery for Time Series Anomaly Detection via Counterfactual Pairing (NeurIPS 2026 · similarity 0.87)
- Detect, Explain, Interpret: An End-to-End Benchmark for Time Series Anomaly Detection, Explainability and Interpretability. (NeurIPS 2026 · similarity 0.87)
- Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection (NeurIPS 2025 · similarity 0.87)
- The Elephant in the Room: Towards A Reliable Time-Series Anomaly Detection Benchmark (NeurIPS 2024 · similarity 0.86)
- On Diffusion Modeling for Anomaly Detection (ICLR 2024 · similarity 0.86)
- A Non-isotropic Time Series Diffusion Model with Moving Average Transitions (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.
