We give a warm welcome to our visiting professor Dr. Peter Jung who has recently joined our group at the AI4EO International Future Lab. He works at the Institute of Telecommunication Systems at the Technical University of Berlin. As guest professor of our Future Lab he will lead together with other core scientists the research activities of the Uncertainties thematic axis.
Dr. Jung has a diploma in Physics from the Humbolt University of Berlin, Germany. In 2007 he received his doctorate degree from the Technical University Berlin with a work titled “Weyl-Heisenberg Representations in Communication Theory”. His research field is described by himself as the interface between applied math and telecommunication engineering.
Next, we share with our readers a short introductory interview we did to him:
Thank you Dr. Jung for accepting the invitation to this short conversation.
Could you please tell us what are your current research topics and interests? could you explain briefly what is it about?
I am working in information and signal processing and focus there on the mathematical treatment of high-dimensional data science problems.
I have an interdisciplinary scope and consider recovery and classification tasks in wireless communication, compressive and computational imaging, group testing, sensor fusion and other physical fields using tools from compressed sensing, low-rank recovery, super-resolution, machine learning and coding and information theory.
For example, future wireless systems and sensor networks will come with high-dimensional inverse problems due to a massive number of nodes, users, antennas and services. Establishing there sporadic and short communication links at low latency in unstable non-stationary situations is a key challenge for e.g. the internet of things and vehicular communication.
Therefore I am interested in artificial intelligence based inference, and recovery approaches, and investigate neurally augmented algorithms for solving inverse problems.
How would you describe the potential synergy between your research and the topics of our lab?
Recovery and classification in Earth Observation is based on sensed data from different sources. To make reliable decisions, careful analysis of uncertainty is crucial. Observations have to be compressed, communicated, processed and fused to gain further knowledge – as in wireless networks. Uncertainty in communication problems may cause erroneous decoding of information messages, hence the analysis of compression and data rates for provably reliable communication are at the core of information theory. Artificial intelligence (AI) has shown superior empirical performance in certain inference tasks and explaining its success on a theoretical level is therefore important in both fields.
What do you think is the biggest challenge of artificial intelligence nowadays?
In many practical problems it is of fundamental importance that we can trust AI architectures beyond empirical evidence. To achieve this a precise mathematical understanding of AI is necessary.
Thank you once more Professor Jung for sharing this and we wish you a nice start in our group activities.
To get further information about Professor Jung and his work we recommend visiting his personal website: http://www.peterjung.eu
