Deep learning

Deep learning has taken the world by storm as the modern day work-horse powering many recent solutions in artificial intelligence, including Chat GPT, Dall-E, and autonomous driving. Deep neural networks are used in many applications in natural language processing and image analysis and have replaced older techniques that require more manual work.

In the early 2000s researchers began to work with deeper models thanks to smarter algorithms and increasing computation power. Deep neural networks can learn internal representations of data, something that previously was labour intensive and required a lot of domain knowledge. In 2012, image analysis was revolutionized by the use of deep convolutional neural networks for image classification. Since then, many advances have been made and today deep learning is an important component for autonomous driving and features in smart phones such as face unlocking. Deep learning has also proved useful in natural language processing. Machine translation and categorization of texts are just a few examples.

At RISE, we have solid expertise in deep learning and many of its application areas. Some examples of related projects include image analysis with the fashion industry, anomaly detection for the automotive industry, privacy and the anonymization of images, and uncertainty calculations for applications in computational fluid dynamics.