2020/10/10 by Adín Ramírez Rivera, Adıń Ramıŕez Rivera, Adil Khan +4
Computer Science · Engineering · Mathematics · Medicine · #Anomaly (physics) #Anomaly Detection Techniques and Applications #Anomaly detection #Artificial Immune Systems Applications #Artificial intelligence #Binary number #Computer science #Data mining #Data-Driven Disease Surveillance #Feature (linguistics) #Feature vector #Mathematics #Outlier #Pattern recognition (psychology) #Physics #Representation (politics) #cs.CV #stat.ML
paper · pdf · doi:10.1109/tnnls.2020.3027667
To appear in IEEE Trans. on Neural Networks and Learning Systems
arxiv created 2020/10/10 · arxiv updated 2020/10/13 · openalex publication_date 2020/10/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06
Anomaly detection suffers from unbalanced data since anomalies are quite rare. Synthetically generated anomalies are a solution to such ill or not fully defined data. However, synthesis requires an expressive representation to guarantee the quality of the generated data. In this article, we propose a two-level hierarchical latent space representation that distills inliers' feature descriptors [through autoencoders (AEs)] into more robust representations based on a variational family of distributions (through a variational AE) for zero-shot anomaly generation. From the learned latent distributions, we select those that lie on the outskirts of the training data as synthetic-outlier generators. Also, we synthesize from them, i.e., generate negative samples without seen them before, to train binary classifiers. We found that the use of the proposed hierarchical structure for feature distillation and fusion creates robust and general representations that allow us to synthesize pseudo outlier samples. Also, in turn, train robust binary classifiers for true outlier detection (without the need for actual outliers during training). We demonstrate the performance of our proposal on several benchmarks for anomaly detection.