MCAKE: Memory-Augmented Autoencoder with Contrastive Learning for Unsupervised Anomaly Detection

Recently, reconstruction-based deep models have gained widespread usage in unsupervised anomaly detection. However, they may overlook some anomalies owing to the over-generalization of neural networks. Several studies have incorporated memory networks to mitigate this problem. Nonetheless, some of them lack an explicit memory updating process, while others rely on data-driven updating methods that are sensitive to initial values and unsuitable for end-to-end training. Additionally, the traditional criterion for detection computed in the high-dimensional input space may collapse as the spike in the deviation score is averaged across numerous dimensions. To address these challenges, we propose MCAKE, a Memory-augmented Contrastive Autoencoder with KNN-Based Extraction. It is designed to highlight the deviation score for anomalies by reconstructing input using fixed normal prototypes recorded in the memory. We explicitly encourage the memory to be autonomously learned and effectively allocated through contrastive learning with multiple positive and multiple negative samples. Furthermore, we introduce a bivariate detection criterion that calculates anomaly scores considering both input and latent space to tackle the collapse. Extensive experiments on 50 datasets across various categories demonstrate the superiority of our approach, with a 2% relative improvement over the previous state-of-the-art models.

• Publication year

  2025

• Venue

  ACM Transactions on Knowledge Discovery from Data

• Publication date

  2025-08-12

• Fields of study

  Computer Science

• Identifiers
  DOI 10.1145/3759460
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• A Light-Weight and Robust Tensor Convolutional Autoencoder for Anomaly Detection

  2024cited by this paper
• Disentangling Tabular Data towards Better One-Class Anomaly Detection

  2024cited by this paper
• Anomaly Detection in Dynamic Graphs: A Comprehensive Survey

  2024cited by this paper
• EntropyStop: Unsupervised Deep Outlier Detection with Loss Entropy

  2024cited by this paper
• Self-supervised enhanced denoising diffusion for anomaly detection

  2024cited by this paper
• Deep Isolation Forest for Anomaly Detection

  2022cited by this paper
• Anomaly Detection for Tabular Data with Internal Contrastive Learning

  2022cited by this paper
• GRELEN: Multivariate Time Series Anomaly Detection from the Perspective of Graph Relational Learning

  2022cited by this paper
• ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions

  2022cited by this paper
• An Improved KNN-Based Efficient Log Anomaly Detection Method with Automatically Labeled Samples

  2021cited by this paper
• RCA: A Deep Collaborative Autoencoder Approach for Anomaly Detection

  2021cited by this paper
• SDFVAE: Static and Dynamic Factorized VAE for Anomaly Detection of Multivariate CDN KPIs

  2021cited by this paper
• Graph Neural Network-Based Anomaly Detection in Multivariate Time Series

  2021cited by this paper
• Learning Memory-Guided Normality for Anomaly Detection

  2020influential reference
• COPOD: Copula-Based Outlier Detection

  2020cited by this paper
• Supervised Contrastive Learning

  2020cited by this paper
• A Simple Framework for Contrastive Learning of Visual Representations

  2020cited by this paper
• PyTorch: An Imperative Style, High-Performance Deep Learning Library

  2019cited by this paper
• Memorizing Normality to Detect Anomaly: Memory-Augmented Deep Autoencoder for Unsupervised Anomaly Detection

  2019influential reference
• Deep Autoencoding Gaussian Mixture Model for Unsupervised Anomaly Detection

  2018influential reference
• Isolation‐based anomaly detection using nearest‐neighbor ensembles

  2018cited by this paper
• Designing Size Consistent Statistics for Accurate Anomaly Detection in Dynamic Networks

  2018cited by this paper
• Memorization Precedes Generation: Learning Unsupervised GANs with Memory Networks

  2018cited by this paper
• Training Adversarial Discriminators for Cross-Channel Abnormal Event Detection in Crowds

  2017cited by this paper
• Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery

  2017cited by this paper
• One-shot Learning with Memory-Augmented Neural Networks

  2016cited by this paper
• End-To-End Memory Networks

  2015cited by this paper
• FaceNet: A unified embedding for face recognition and clustering

  2015cited by this paper
• Neural Machine Translation by Jointly Learning to Align and Translate

  2014cited by this paper
• Memory Networks

  2014cited by this paper
• Auto-Encoding Variational Bayes

  2013cited by this paper
• Credit Card Fraud Detection Using Hidden Markov Model

  2012cited by this paper
• Histogram-based Outlier Score (HBOS): A fast Unsupervised Anomaly Detection Algorithm

  2012cited by this paper
• Isolation Forest

  2008cited by this paper
• Dimensionality Reduction by Learning an Invariant Mapping

  2006cited by this paper
• A Novel Anomaly Detection Scheme Based on Principal Component Classifier

  2003cited by this paper
• Estimating the Support of a High-Dimensional Distribution

  2001cited by this paper
• LOF: identifying density-based local outliers

  2000cited by this paper
• Efficient algorithms for mining outliers from large data sets

  2000cited by this paper
• Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence Fast Anomaly Detection for Streaming Data ∗

  year unknowncited by this paper

• Gated Memory-Guided Multi-scale spatio–temporal–spectral feature fusion network for unsupervised Internet of Things time series anomaly detection

  2026cites this paper
• A Method for Suppressing Over-Reconstruction in Diffusion Models for Video Anomaly Detection

  2025cites this paper