Multimodal AI security systems promise smarter, more private smart homes
A systematic review of 87 studies identifies fragmentation across multimodal fusion, adaptive learning, edge inference, privacy and explainability as the ...
Tag: Anomaly Detection
From Pixels to Compressed Streams: A Review Maps the Evolving Field of Anomaly Detection
A new review in Knowledge and Information Systems synthesizes a decade of anomaly detection research across images, videos, and compressed ...
New AI Method OSCAR Reveals Hidden Brain Network Reorganization in Tasks and Psychosis
A new anomaly-detection method called OSCAR learns each brain region's typical whole-brain connectivity profile and flags multivariate deviations during tasks ...
Survey Maps Two Decades of Intelligent Video Surveillance Systems
A systematic review in the Journal of Big Data synthesizes research from 2000 to 2025 on intelligent video surveillance, proposing ...
Hybrid encryption protects federated updates in Medical-MNIST test
A Paillier-CKKS framework encrypted federated-learning updates and reached 94.67% average accuracy on Medical-MNIST during simulated reverse-engineering attacks.
AI-Powered 5G Framework Slashes Mobile Energy Use While Boosting Security
Researchers have built a three-layer AI framework for 5G mobile learning that cuts device energy use by up to 27 ...
AI Watches the Wind and Sun: Transformer Model Learns to Spot Grid Anomalies
Researchers in India built a transformer-based deep learning system to detect anomalies in solar-wind power plants, achieving strong reconstruction accuracy ...
Federated learning brings privacy-first anomaly detection to smart grids
Smart grids promise a cleaner, more responsive electricity system, but they also carry a hidden vulnerability: every smart meter is ...
AI Forensic Accounting Screens Palestinian Banks for Hidden Instability Risk
Machine learning and SHAP explainability reveal conditional disclosure-stability patterns across seven Palestinian banks, offering supervisors an early-warning triage tool.
Blockchain Meets Machine Learning to Guard Wireless Sensor Networks Against Attackers
A new framework fuses an unsupervised machine learning ensemble with a lightweight blockchain to detect and confirm intrusions in wireless ...
AI Learns to Spot Failing EV Batteries Before They Catch Fire
Researchers have developed an unsupervised deep learning model called GSA-VAE that detects faults in electric vehicle batteries by fusing causal ...
Dual-encoder AI framework sharpens anomaly detection in multivariate time series
Researchers at East China Jiaotong University have developed DE-SVDD, a dual-encoder framework that separates reconstruction learning from one-class boundary modeling ...
Self-Attention AI Learns to Predict CNC Tool Failure Before It Happens
A compact deep learning model that mathematically separates normal machining signals from anomalous disturbances has set new benchmarks for predicting ...
AI Architecture Joins Load Forecasting, Anomaly Detection, and Predictive Energy Management
The distinction between the measured dataset and the synthetic temporal extension deserves particular attention, because it reflects a broader methodological ...
AI Watchdog That Learns Normal Water and Spots Contamination in Real Time
A new AI-IoT framework combining an LSTM autoencoder with Random Forest regression achieves near-perfect anomaly detection and accurate water quality ...
Deep Learning AI Detects and Repairs Hacker Attacks on Brain Signals in Real Time
A new deep learning framework can detect six types of cyber-attacks on EEG signals and reconstruct the corrupted brain data ...
Deep Learning’s Anomaly Detection Supremacy May Be an Illusion, Study Finds
A reproducibility study of 18 influential papers finds that data leakage, silently modified datasets, and underspecified baselines have inflated claims ...
AI Learns to Explain Why Data Goes Wrong, One Feature at a Time
Researchers at the University of Calabria have developed RLAX, a reinforcement learning framework that identifies compact, interpretable feature subspaces explaining ...
Deep Learning Meets Blockchain: VoteChain Slashes Energy Use in Financial Consensus
Researchers have developed VoteChain, a voting-based blockchain consensus framework that embeds a lightweight CNN-LSTM anomaly detector to cut energy use ...
A Classic Statistics Test Gets a Streaming Makeover to Catch Anomalies in Real Time
Researchers in Seoul have rebuilt the classic extreme studentized deviate test into an online, self-tuning algorithm that detects anomalies in ...
POPULAR NEWS
-
Two C3 Gene Variants Reveal Functional Independence Between Extracellular and Intracellular Forms of Complement C3
-
Nearly Half of Stroke Patients in Africa and Asia Show Low HDL Cholesterol
-
German heat-health plans rarely evaluated due to political and structural barriers, study finds
-
Carb Quality Beats Glycemic Index in Sperm DNA Study