This presentation will explore the use of deep leaning technology in machine vision applications. It will specifically explore the effectiveness of CPU-based deep learning inference including the benefits and considerations to choosing that approach.

In this presentation, we will review classical deep learning-based tools such as classifiers, segmenters (semantic, as well as unsupervised), and objects detectors. Focusing mainly of their advantages and drawbacks, we will explain how these tools meet typical requirements of machine vision applications, and how to choose between them.

An overview of utilising a method of AI/Deep Learning from initial challenge to solution deployment. The session will look at the types of challenges that the technology can help solve, methods of structuring the teaching process and then how to deploy and continually evolve and optimise.

In this video, you can see a detailed demonstration of MVTec’s anomaly detection based on deep learning. MVTec’s deep-learning-based anomaly detection significantly facilitates the automated surface inspection for, e.g., detection and segmentation of defects. The technology is able to unerringly and independently localize deviations, i.e., defects of any type, on subsequent images. You only need a low number of high-quality images for training because defects of varying appearance can be detected without any previous knowledge or any preceding labeling efforts. Training a new network can mostly be done in a matter of seconds, allowing users to perform many iterations to fine-tune their application without sacrificing a lot of precious time.

NET introduces latest smart vision technologies for a faster ´going smart´. The discussed, freely configurable smart vision solutions reduce complexity and allow a faster time-to-market. Concrete applications examples are given.