DeepLTK Anomaly Detection Addon
The DeepLTK Anomaly Detection Addon enables high-performance, unsupervised visual anomaly detection, ideal for defect detection, industrial inspection, and outlier identification.

Key Features
•High-Speed Training: <1 min (GPU), ~5 min (CPU)
•Fast Inference: <10 ms (GPU), ~300 ms (CPU)
•High Accuracy: >95% on 80+ datasets
•Unsupervised: Learns from normal samples only - no annotations needed
•Few-Shot Learning: 100s of good images can achieve near-perfect accuracy
•Versatile: Ideal for manufacturing, medical imaging, and predictive maintenance

Examples: https://github.com/ngenehub/deepltk_examples/tree/main/3_Image_Processing/36_Anomaly_Detection
Tutorials: https://www.ngene.co/post/deepltk-tutorial-3-6-visual-anomaly-detection