IDS NXT Lighthouse Training Software Helps To Make Deep Learning Accessible To Everyone

In order to implement AI vision solutions, specialist knowledge, development effort and investment in computer hardware and storage technology are usually required.

With cloud computing and specialized training services such as IDS NXT lighthouse, the process becomes much easier: neural networks can be trained individually and at the push of a button, without the need to set 978YIDS NXT ocean, the training software helps to make deep learning accessible to everyone. With the release of IDS NXT lighthouse version 1.1, additional functions are now available.

Powerful inference cameras

IDS NXT lighthouse is an easy to use training software for creating neural networks (IDS Imaging Development Systems GmbH) The training of neural networks in the web application requires three basic steps: to upload individual sample images, to label the images (e.g. "good" / "bad") and then to start the fully automatic training.

The generated network can be executed directly on IDS NXT industrial cameras. This turns them into powerful inference cameras that can solve completely new image processing tasks thanks to artificial intelligence. They have learned which image features are important and are able to apply this knowledge to new image data. This is an enormous advantage compared to classical programming, especially in applications with highly varying objects.

Different sensors and protection classes

All components – camera hardware with AI core and different sensors and protection classes, software, infrastructure, knowledge and support – are developed and provided by a single supplier. This holistic approach makes IDS NXT ocean a unique solution in the vision market.

The IDS NXT lighthouse training software for neural networks is now available as version 1.1. The company has further improved both usability and the range of functions, for example by redesigning the detail view or extending filter functions. The next version is already in the works – then, in addition to classification tasks, object detection will also be feasible.