This project implements an unsupervised anomaly detection system for industrial inspection using deep learning. The model identifies defects in manufacturing products by analyzing images and detecting deviations from normal patterns.

• Uses Anomalib, a specialized library for anomaly detection.
• Implements PatchCore, a state-of-the-art anomaly detection model.
• Supports MVTec AD, an industrial anomaly detection dataset.
• Visualizes anomaly heatmaps and masks for better interpretability.
• Runs efficiently on GPU (T4/P100) or TPU (v3-8).

• MVTec AD: A widely used benchmark dataset for anomaly detection.
• Categories: Bottles, Cables, Capsules, Metal Nuts, etc.
• Each category contains normal and anomalous images for training and evaluation.

Ensure you have Python 3.8+ and install the required dependencies:

pip install anomalib
pip install wandb

• PatchCore is used due to its efficiency and high detection accuracy.
• Anomaly heatmaps highlight defective areas in the images.
• Resizing and preprocessing ensure compatibility with the model.

• Training is performed on normal images to learn the standard patterns.
• Inference is run on test images to classify and localize anomalies.
• Visualization helps interpret the results effectively.

• Generates heatmaps and binary masks for defect localization.
• Achieves high precision and recall on the MVTec AD dataset.
• Can be fine-tuned for custom industrial datasets.

• Black images in visualization? Ensure proper image normalization.
• ModuleNotFoundError? Check if anomalib is correctly installed.
• Dataset not found? Verify the dataset path in your notebook.

• Implement real-time anomaly detection.
• Explore self-supervised learning for improved performance.
• Deploy the model for industrial automation.

• Anomalib Documentation
• MVTec AD Dataset