anomaly-detection-algorithm

[ICLR 2026] The implementation of the paper Foundation Visual Encoders Are Secretly Few-Shot Anomaly Detectors

Semi-supervised anomaly detection method

[CIKM 2021] A PyTorch implementation of "ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning".

Anomaly detection method that incorporates multi-scale features to sparse coding

Detects anomalous resting heart rate from smartwatch data.

This project provides a time series anomaly detection algorithm based on the dynamic threshold generation model.

Artificial Immune Systems Package (AISP) is an open-source Python library that features bio-inspired algorithms based on artificial immune systems for machine learning, pattern recognition, anomaly detection, and optimization tasks.

OCR to detect and recognize dot-matrix text written with inkjet-printed on medical PVC bag

Several examples of anomaly detection algorithms for time series data.

An official source code for paper "Normality Learning-based Graph Anomaly Detection via Multi-Scale Contrastive Learning", accepted by ACM MM 2023.

The paper "Deep Graph Level Anomaly Detection with Contrastive Learning" has been accepted by Scientific Reports Journal.

Uses LSTM-based autoencoders to detect abnormal resting heart rate during the coronavirus (SARS-CoV-2) infectious period using the wearables data.

DecompositionUMAP: A multi-scale framework for pattern analysis and anomoly detection

Kernel-Level Energy Profiling & Anomaly Detection for Android

Anomaly detection algorithm for time series based on the dynamic threshold generation model

UPIQAL is a 100% automated Full-Reference Image Quality Assessment (FR-IQA) framework. It synthesizes deep feature statistics, probabilistic uncertainty, and spatial heuristics to replace human MOS and output spatially localized diagnostic heatmaps for visual artifacts.

Benchmarking anomaly detection on air-quality time series using a tiny autoencoder and classical methods. Includes reproducible pipeline, strict time-aware evaluation, and analysis of thresholding strategies and model performance gaps.

an end to end anomaly intrusion base on deep learn

SCR# 3056.0: WGAN-based Digital Twins for Anomaly Detection