Deep learning in MERLIC
In machine vision, deep learning enables an increasing number of applications that weren’t possible before. Moreover, the performance of existing applications can be significantly improved. Especially in the field of data classification, deep learning is a very efficient technology.
With MERLIC from MVTec, the use of these deep learning methods is very easy to implement. The pre-defined deep learning tools can be integrated into the tool flow of the machine vision application via drag-and-drop and executed.
Deep learning methods in MERLIC
With the Deep Learning Tool, you can easily label your data and train your models thanks to the intuitive user interface – all without programming skills. This data can be seamlessly integrated into MERLIC to run deep learning-based processes.
Certain requirements are necessary to be able to use the above-mentioned technologies. The following table shows what is needed to use deep learning in MERLIC.
| Technology | Model training in DLT required? | MERLIC Deep-Learning-AddOn required? |
| Classification | ✔ | ✔ |
| Anomaly Detection | X (For this technology, training of the model can be done directly in MERLIC.)
| ✔ |
| Anomaly Detection in the Global Context | ✔ | ✔ |
Object Detection Instance Segmentation | ✔ | ✔ |
| Segmentation | ✔ | ✔ |
| Deep OCR* | X (For this technology, training of the model can be done directly in MERLIC.) | X |
| Deep Counting* | X (For this technology, training of the model can be done directly in MERLIC.) | X |
*This technology is included in the basic MERLIC package.