Meet the Helmholtz AI local unit for information @ FZJ
‘Humans of Helmholtz AI’ #2: Helmholtz AI @ FZJ
At Helmholtz AI we aim to push boundaries in terms of artificial intelligence and machine learning, but this still requires some human-powered efforts. Let us introduce the ‘humans’ who make Helmholtz AI possible!
This month, we spoke with the eight humans who compose the Helmholtz AI local unit for research field information at Forschungszentrum Jülich (FZJ).
Morris Riedel, head of the FZJ local unit and member of the Helmholtz AI steering board, contributes to the platform’s mission with ‘extensive parallel and scalable algorithm & technology know-how and embedding Helmholtz AI activities into the larger AI strategy of FZJ and its partners’. This task is also supported by Susanne Wenzel, the local coordinator, who is in charge of internal communication and regular exchange with colleagues from the whole Helmholtz AI platform. ‘I make sure that everything runs smoothly and the next steps are implemented’, she says.
Timo Dickscheid is head of “Big Data Analytics” at the Institute for Neuroscience and Medicine (INM-1), and leads the new research group 'AI methods for building ultrahigh resolution human brain models' of Jülich’s Helmholtz AI local unit. 'We develop new deep learning methods for analysis and 3D reconstruction of diverse microscopic image modalities, and interact closely with researchers in brain-inspired AI to make use of our models of the microstructural organization of the human brain', he explains. Christian Schiffer and Eric Upschulte are PhD candidates in the team. They are both developing deep learning algorithms that contribute to the creation of a microstructural reference atlas of the human brain.
As for research and high level support at Jülich Supercomputing Center (JSC), there is Jenia Jitsev, scientific lead for machine / deep learning, responsible for defining general research directions to be followed within the Helmholtz AI local unit at JSC. These directions are aligned with the agenda of the high level support team, other research groups at JSC and with partners at INM-1. Jenia also leads the cross-sectional team deep learning that works closely with the consultants to push forward generic large-scale deep learning methods applicable across different tasks and domains. As Helmholtz AI consultants, Mehdi Cherti and Alexandre Strube are members of JSC’s high level support team, providing support to the Helmholtz AI research community for their scientific projects, performing sustainable software engineering and also conducting research on machine learning / deep learning taking advantage of the supercomputers available at Jülich.
Helmholtz AI: research expertise, interdisciplinary exchange, unique network
When asked what they enjoy most about being part of Helmholtz AI, answers are quite similar even for those like Mehdi who joined the platform quite recently. He is ‘impressed by the diverse range of scientific domains and topics that are included in Helmholtz AI’, which for Christian and Eric translates into great ‘opportunities for interdisciplinary exchange, and a variety of interesting challenges’.
Timo describes Helmholtz AI as ‘a great opportunity for the researchers inside Helmholtz to team up on methods development and data exchange, and to engage with the international AI community by exchanging knowledge, people and data. This is an ‘effective approach’ that can also lead to ‘international visibility’, according to Morris.
On the other side, Susanne, who was involved in creating the Helmholtz initial concept for the platform almost from the beginning, also highlights ‘the spirit of bringing together scientists from the whole association and Helmholtz' broad range of research domains’ as a very special achievement almost unique in Germany. Jenia also likes ‘the idea to create a large, open collaborative research network dedicated to advanced machine learning methods and their application in science. (...) If we manage to use this potential properly, enriching it with critical mass of collaborative machine learning expertise and software engineering, great things lie ahead. As Alexandre puts it, ‘It's nice to be riding the wave!’
Life beyond Helmholtz AI
Most of the local unit members have a common background in computer/data science, but their hobbies outside Helmholtz AI are much more diverse.
For Alexandre, who calls himself a ‘full-fledged trekkie’, it is all about riding motorcycles and bicycles, sailing, as well as making things with sensors, electronics, arduino/raspberry pi and so on. Susanne also enjoys engaging in DIY projects, and technical toys like Lego technics or mindstorms, together with her kids.
There are also some musicians in the team; Timo plays keyboards in jazz and soul bands, and Eric loves playing the drums and guitar. Mehdi is into watching movies, playing football with friends, and listening to music. Christian also enjoys books, programming and videogames.
Jenia also likes good books, music and movies, but also travelling with his friends and his daughter around the world, playing table tennis and basketball, and engaging in social awareness initiatives like March for Science. He is inspired by ‘people who do not give up on the idea of studying and improving the world we all live in a bit despite all the despair, struggles and setbacks on the way there’.
On the nature side, Morris enjoys running outdoors or spending time with family and friends. He remarks that people would be surprised to know that Iceland (where he spent a few years of his life) is ‘not always as cold as people think and that you can nicely relax after a tough working day in a nice hotspring deep in nature, which I quite often do’.
*Copyright picture: Forschungszentrum Jülich / Ralf-Uwe Limbach