Helmholtz AI

Discover Helmholtz AI: Democratising AI across the Helmholtz Association

Helmholtz AI is one of five platforms initiated by the Helmholtz Information and Data Science Incubator. Its main goal is to become a driver for applied artificial intelligence (AI) through the development and distribution of AI methods across all Helmholtz centres, combining AI-based analytics effectively with Helmholtz' unique research questions and datasets. By implementing a Helmholtz-wide network for applied AI, Helmholtz AI builds on the current AI strengths in the Helmholtz Association and brings together scientists from all centres, provides support in their needs and thereby fosters transdisciplinary and ground-breaking research. Democratising access to AI maximises research impact, enabling Helmholtz AI to potentially reach an international leadership position in the field.

Helmholtz AI is organised into one central and five local units operated at different Helmholtz centres and supporting the six research fields in the association. The local unit for the Research field “Information” is established at Forschungszentrum Jülich and implemented with a research group at the Institute of Neuroscience and Medicine (INM-1) and an AI Consultant team at the Jülich Supercomputing Centre (JSC). It works at the intersection of AI, high-performance computing (HPC), and neuroscience, with a focus on methods for biomedical image analysis and the training and operation of large-scale AI models on supercomputing clusters.

The local unit offers direct support for researchers through vouchers, Helmholtz AI projects, and consultancy, so do not hesitate to get in touch.

Contact:

Member of Helmholtz AI Steering Board: Timo Dickscheid (FZJ)
Head of Helmholtz AI Consultant team at JSC: Stefan Kesselheim (FZJ)
Scientific Coordination: Susanne Wenzel (FZJ)

Find out more

Helmholtz AI for Research field Information in detail

Artificial Intelligence helps researchers decode the organisation of the human brain

Copyright: Forschungszentrum Jülich / Ralf-Uwe Limbach, Markus Axer, Kai Kiwitz, Eric Upschulte, Timo Dickscheid

The Helmholtz AI research group at INM-1 develops AI methods for analysing the microstructural organisation of the human brain, especially including deep learning methods for object detection and texture analysis with limited training data. Driven by high-throughput imaging and the ambition to quickly transfer knowledge across tasks and data domains, the implementation of AI methods on HPC clusters is a key aspect. The group brings a network of strategic partnerships at the intersection of neuroscience, AI, and HPC, including the Helmholtz International Lab HIBALL with McGill University and Mila (Montreal), the University of Iceland, and a range of European labs working in this field in the frame of the Human Brain Project.

Contact: Christian Schiffer (FZJ), Timo Dickscheid (FZJ)

Learn more
Helmholtz AI consultant teams ensure translation of expertise into the immediate research domain of their respective units.

The Helmholtz AI consultant team for research field Information is associated with exciting research domains such as neuroscience, quantum computing, and material science. The consultants support scientists with their AI competence in all phases of their research projects; from sketch through implementation to communication. A specific focus of the team is the use of Jülich’s supercomputing systems for large-scale AI applications, especially to stimulate interest in large-scale pre-training, self-supervised methods, generative models, and the use of physical priors.

Contact: Stefan Kesselheim (FZJ)

Learn more
Empowering Helmholtz AI with top-tier GPU resources & expert guidance at FZJ & KIT

Access to computing resources is key for the Helmholtz AI community to accelerate innovative AI applications. HAICORE provide easy and low-barrier access to graphics processing units (GPUs) for the entire AI community within the Helmholtz Association. Dedicated compute clusters are installed and managed at FZJ and KIT. Together, the AI consultant teams at KIT and FZJ support the HAICORE users by providing testing and benchmarking the infrastructure, preparing and maintaining AI-related software installations and with first level support.

Contact: Stefan Kesselheim (FZJ), Markus Götz (KIT)

Learn more

Test

Test
Test

Helmholtz AI

Helmholtz AI Research group for Research field Information

Helmholtz AI Consultant Team for Research field Information

HAICORE: Computing resources for the Helmholtz AI community