ANN4EUROPE: ANNs for efficient, robust and interpretable reconstruction in Physics and beyond



Our joint research focuses on two highly complex experiments: the LHCb experiment in high-energy physics and the CBM experiment in heavy-ion physics. The LHCb experiment is already in operation, and the CBM experiment is in planning at the world's largest accelerator facilities, LHC (CERN, Geneva, Switzerland) and FAIR (GSI, Darmstadt, Germany). These experiments are the most difficult, not only in terms of the extreme complexity of the physics itself, when it comes to exploring the asymmetry between matter and anti-matter at LHCb, or the study of the superdense state of matter, as in the center of neutron stars, at CBM. These experiments also pose increasing challenges to the fast processing and analysis of the huge amounts of experimental measurements at more than 10 million collisions of ultrarelativistic particle beams per second, with data volumes up to 500 Tbit per second. Given the increasing fragmentation of computer architectures and their specialization in AI (Artificial Intelligence) tasks, the successful solution of fast data processing problems requires the development of new algorithms and techniques based on AI methods.

We will develop deep neural networks that enable low-cost, energy-efficient data processing in a variety of high-energy and heavy-ion physics experiments.   Together, our teams will be able to use AI to maximize the performance of tomorrow's hybrid computing architectures, improve computational and energy efficiency by orders of magnitude, and support future fundamental research. Because the volume and complexity of the data we are dealing with in our experiments exceed the most demanding industrial applications, we can expect the success of this project to spread to other areas where real-time pattern recognition with AI is required. Examples include self-driving cars, smart cities, real-time detection of anomalies in medical procedures or industrial processes, and many others.

Project leader


Ivan Kisel

External Partners

Prof. Dr. Vladimir Vava Gligorov

Laboratoire de Physique Nucléaire et de Hautes Energies (LPNHE), Paris