🤖 AI Summary
Time-series anomaly detection faces three key challenges: limited contextual modeling, inadequate representation of normal patterns, and inaccurate evaluation metrics. To address these, we propose SimAD—a lightweight discrepancy modeling framework. SimAD introduces an EmbedPatch encoder that jointly models local and global temporal features, a ContrastFusion module to enhance separability between normal and anomalous distributions, and theoretically grounded robust evaluation metrics—UAff (Unsupervised Affinity) and NAff (Normalized Affinity). Evaluated on seven standard benchmarks, SimAD consistently outperforms state-of-the-art methods, achieving absolute improvements of 19.85% in F1-score, 77.79% in NAff-F1, and 9.69% in AUC. All code and pretrained models are publicly released.
📝 Abstract
Despite the prevalence of reconstruction-based deep learning methods, time series anomaly detection remains challenging. Existing approaches often struggle with limited temporal contexts, inadequate representation of normal patterns, and flawed evaluation metrics, hindering their effectiveness in identifying aberrant behavior. To address these issues, we introduce $ extbf{{SimAD}}$, a $ extbf{{Sim}}$ple dissimilarity-based approach for time series $ extbf{{A}}$nomaly $ extbf{{D}}$etection. SimAD incorporates an advanced feature extractor adept at processing extended temporal windows, utilizes the EmbedPatch encoder to integrate normal behavioral patterns comprehensively, and introduces an innovative ContrastFusion module designed to accentuate distributional divergences between normal and abnormal data, thereby enhancing the robustness of anomaly discrimination. Additionally, we propose two robust evaluation metrics, UAff and NAff, addressing the limitations of existing metrics and demonstrating their reliability through theoretical and experimental analyses. Experiments across $ extbf{seven}$ diverse time series datasets demonstrate SimAD's superior performance compared to state-of-the-art methods, achieving relative improvements of $ extbf{19.85%}$ on F1, $ extbf{4.44%}$ on Aff-F1, $ extbf{77.79%}$ on NAff-F1, and $ extbf{9.69%}$ on AUC on six multivariate datasets. Code and pre-trained models are available at https://github.com/EmorZz1G/SimAD.