AI fraud detection AI Unmasks Financial Fraud: How Dual-Path Graph Filtering Boosts Detection Accuracy Explore DPF-GFD, an AI model using dual-path graph filtering to combat financial fraud by overcoming relation camouflage, heterophily, and data imbalance.
IEC-61850 Securing Digital Substations: Real-Time Anomaly Detection in GOOSE Networks with Unsupervised AI Explore how unsupervised temporal AI models provide real-time intrusion detection for IEC-61850 GOOSE networks in digital substations, overcoming latency and data challenges.
anomaly detection PASTA: Revolutionizing Anomaly Detection with Weakly Supervised Vision Transformers Discover PASTA, an innovative weakly supervised AI pipeline for precise target and anomaly segmentation in industrial and agricultural settings, leveraging Vision Transformers to reduce training time and enhance operational efficiency.
TinyML Autonomous Anomaly Detection: Bringing TinyML Intelligence to Resource-Constrained Devices Explore Z-Score based TinyML for fully autonomous, on-device anomaly detection using power side-channel data. Learn how AI on microcontrollers enhances efficiency, privacy, and reliability without cloud dependency.
cybersecurity Fortifying Cybersecurity: How Generative AI Combats Unknown Threats in Intrusion Detection Explore GMA-SAWGAN-GP, a cutting-edge generative AI framework that enhances Intrusion Detection Systems (IDS) against known and novel cyberattacks. Discover how advanced data augmentation improves network security and operational resilience.
Cross-Domain Learning Unlocking Cross-Domain Intelligence: How AI Finds Universal Laws for Robust Solutions Explore Importance Inversion Transfer (IIT), a breakthrough AI framework that uncovers universal organizational principles across diverse systems to enhance anomaly detection and AI-powered analog circuit design.
IoT security Boosting IoT Security: Explainable AI and Decision Trees for Anomaly Detection Discover a new AI framework combining optimized Decision Trees with Explainable AI (SHAP, Morris) for real-time, highly accurate, and transparent IoT anomaly detection on edge devices.
