Main Content

Edge AI

Artificial intelligence as a research topic is finding more and more areas of application due to immense progress in recent decades.
In the past, AI systems were primarily based in data centres where collected data were pooled, and large computing capacities were available for training and inference (cloud computing).

Edge computing breaks with this principle and takes the approach that most calculations are carried out as close as possible to the data source (e.g. on a smartphone). Due to the significantly reduced computing power, software architectures must be optimised, and specialised hardware must be developed.

In this seminar, we will examine and compare current algorithms and technologies that enable AI applications on edge devices. These include, for example, architectures optimised for mobile devices, quantisation and compression of deep neural networks.