| Tutorial, HALCON, Deep Learning

Global context anomaly detection in MVTec HALCON

In this tutorial, you will learn about using deep-learning-based global context anomaly detection in MVTec HALCON.

With this technology, you can detect structural and logical anomalies in your images. In this video, we will go through the steps necessary to train and deploy a model. First, we talk about the data preparation, including possible steps to improve the images and preprocessing parameters. Next, we train the model. After the training, we define some thresholds that we need to decide which pixels and images represent anomalies. Lastly, we evaluate the model, and then deploy the model during inference.

  • 0:00 Introduction
  • 0:45 Data preparation
  • 2:01 Training
  • 2:59 Evaluation
  • 4:57 Inference
  • 6:00 Comparison of global and local network

Please note: Once you watch the video, data will be transmitted to Youtube/Google. For more information, see Google Privacy.