The Fraunhofer-Gesellschaft (www.fraunhofer.com) currently operates 76 institutes and research institutions throughout Germany and is the world’s leading applied research organization. Around 30 000 employees work with an annual research budget of 2.9 billion euros.
The »Broadband & Broadcast« department is active in the areas of mobile communications, embedded ML / AI, Internet-of-Things and automotive communication systems. We take new concepts and algorithms in the fields of communications and digital signal processing (e.g. machine learning) from theory, implement them and test them in simulations and in prototypes in our labs and in the field.
You are interested in the field of communication systems and would like to develop further in the field of deep neural networks?
Then have a look at our offer!
Designing neural networks (NNs) involves optimizing many parameters, which make the search space very large. These parameters are for example: The number of layers in NN, the types of each layer, the size of each layer (filters) and the quantization details of each layer. Often, it is not feasible to optimize these parameters by hand or even by brute force. Therefore, neural architecture search strategies and algorithms, have to be deployed to optimize all the decisions affecting the design of NNs.
In this work, existing solutions of NAS should be identified, implemented and analyzed to one another.