The Hessian Minister of Science Timon Gremmels (SPD) visited hessian.AI accompanied by Bijan Kaffenberger (SPD), a member of the state parliament. The Minister was impressed by the research projects, which also play an outstanding role in the federal government’s Excellence Strategy competition. The innovation potential of the AI start-ups closely associated with hessian.AI also met with keen interest.
After the welcome address by the President of TU Darmstadt, Professor Tanja Brühl, the co-directors of hessian.AI Professor Mira Mezini and Professor Kristian Kersting, provided insights into the diverse work of the center.
Cutting-edge research: RAI and TAM cluster projects
Kristian Kersting and Marcus Rohrbach presented the projects “Reasonable Artificial Intelligence (RAI)” and “The Adaptive Mind (TAM)”. The planned Cluster of Excellence RAI aims to develop a new generation of AI systems with the advantages of reasonable use of resources, data protection and continuous improvement. “The Adaptive Mind (TAM)” examines human behavior under changing external conditions. TAM creates new approaches to understanding and computationally modeling human perception, thinking, decision-making, action and learning.
Exceptional infrastructure
Such complex research projects require high computing power. The unique infrastructure of hessian.AI with the supercomputer fortytwo, the AI Innovation Lab and the AI Service Center provides the ideal framework conditions for the further development and application of AI systems. Florian Kieser, Head of the AI Service Center, presented the three building blocks of the infrastructure and the wide range of further training courses offered by hessian.AI.
Dynamic AI innovation ecosystem
Timon Gremmels and Bijan Kaffenberger were able to see the great practical benefits of AI systems for themselves in the subsequent demonstrations:
Representing Hessen’s dynamic AI startup system, the founders of the startups Bird Mapper, Energy Roboter and etalytics, who are closely associated with hessian.AI, showed their AI innovations.
Bird Mapper is an AI-based solution for accurately recording bird populations – a contribution to preserving biodiversity. Founder and TU Darmstadt student Marc Neumann developed his idea to market maturity with the support of funding programs from hessian.AI (AI Startup Rising) and TU Darmstadt (Highest). The young company etalytics, founded in 2020, is also committed to sustainability: with its AI-supported software etaONE®, the startup reduces the energy consumption of IT systems. The startup Energy Robotics, which provides autonomous robots for special maintenance work on industrial plants, is already being used successfully around the world.
In addition to the young entrepreneurs, young researchers also presented two pioneering projects: The Llava Guard project, which was created as part of the RAI project, offers a method with which image content in large data sets or from image generators can be filtered, evaluated and offensive content suppressed for the first time using vision-language models.
Minister Gremmels also tested an interactive application of the Occiglot language model using the Furhat language robot: the project aims to create a coherent language modeling system that takes into account all 24 official languages of the European Union as well as other unofficial and regional languages.
Photo: TU Darmstadt/Klaus Mai
Prof. Dr. Carsten Binnig is Professor of Data Management at TU Darmstadt and Head of the Department of Data and AI Systems. He is also a founding member of hessian.AI and head of the research area Systemic AI for Decision Support at the Darmstadt site of the German Research Center for Artificial Intelligence DFKI.
Carsten Binning and his team are working on the development of a new generation of AI-centric database systems that are self-optimizing and can also be easily operated by non-IT experts. The expert in artificial intelligence (AI) and data management has now been awarded a LOEWE top professorship at TU Darmstadt. The LOEWE professorship will be funded over five years with around two million euros from LOEWE funds.
State program supports application for “Reasonable Artificial Intelligence” cluster of excellence
By awarding the LOEWE professorship, the state of Hessen is supporting the “Reasonable Artificial Intelligence (RAI)” research project as part of the Excellence Strategy of the German federal and state governments. The planned Cluster of Excellence under the leadership of TU Darmstadt in collaboration with the universities of Frankfurt, Bonn and Würzburg aims to develop the next generation of AI: AI systems that learn with a “reasonable” amount of resources based on “reasonable” data quality and “reasonable” data protection.
Photo: Katrin Binner
About Subarnaduti Paul
Subarnaduti Paul began his scientific career in Germany about five years ago with a Master’s degree at the Technical University of Darmstadt. After completing his studies, which initially focused on electronic systems, Paul moved more and more toards artificial intelligence, inspired by an internship and a subsequent master’s thesis at Bosch.
The core of his research: continuous learning
Paul’s previous work has focused on distributed systems, particularly federated systems, in which multiple clients learn data locally and then send this information to a global server.
His current research interest is in continuous learning. He explains how this concept aims to teach machines to memorize information in a similar way to humans: “We learn to read and write as children and remember these basic skills decades later.” He compares this to the challenge for machines when they are sequentially fed new data and often forget what they have previously learned – a phenomenon known as catastrophic forgetting.
Recently, he has been focusing on so-called base models, for example for image recognition, and on the integration of continuous learning in an environment with limited memory. “I’m working on how we can incrementally train the Vision Transformer on a more economical model,” explains Paul. “It’s about making models more efficient in sequential learning by focusing on the essential parts of the data,” he explains. This research could help to reduce the memory requirements and computing power for machine learning.
Subarnaduti Paul
Replay buffer as memory
One of the biggest challenges is the development of interpretation methods that make it possible to identify and use the most relevant parts of a data set. Paul illustrates this using the example of a photo: “If we have a picture of a dog against the background of a park, the model has to decide whether it is a dog or another animal. The challenge is to select and save only the relevant part of the image – in this case the dog.
This task requires not only a deep understanding of how AI models work, but also innovative approaches to deal efficiently with limited storage space. Paul’s goal is to design replay buffers that store only a fraction of the data, but contain enough information to effectively support continuous learning. These buffers serve as a type of memory storage that allows AI models to learn from past experiences while minimizing the problem of catastrophic forgetting.
The balance between memory efficiency and retaining relevant information is a balancing act and is the focus of Paul’s research.
Continuous learning could make AI fit for global applications
For Paul, hessian.AI offers an ideal platform for his research. “It’s great to have so many PhD students from different fields under one roof. The opportunity to collaborate and the access to resources such as large GPU clusters are crucial for my work,” Paul emphasizes.
Paul sees great potential for social progress in his field of research. As an example, he cites the gradual introduction of vaccines for different age groups, which follows the principle of continuous learning.
“Through continuous learning, we could reduce distortions in data and models and achieve greater generalization across different population groups,” he explains. This also applies to cases where, for example, a language model was initially trained with European languages and is now to be extended to other languages.
In the future, Paul therefore plans to extend his research to socially relevant applications and also to advocate the development of standardized benchmarks in the field of continuous learning. In this way, he hopes to significantly advance this young field of research.
About Aryaman Reddi
Aryaman Reddi is a PhD student working on multi-agent reinforcement learning and game theory. He began his academic career at the prestigious University of Cambridge, where he completed both his Bachelor’s and Master’s degrees in Information and Computer Engineering. During his studies, Reddi became interested in machine learning and mathematics. His Master’s thesis entitled “Deep Q-Learning for Congruent Non-Dominated Game Strategies” laid the foundation for his current research.
After his studies, Reddi gained practical experience at ARM, where he worked in the machine learning research department. His work focused on the optimization of neural networks for applications such as facial recognition on mobile devices.
Today, Reddi continues his research at the Technical University of Darmstadt, where he is a member of Professor Carlo D’Eramo’s LiteRL team. There, he is working intensively on the development of multi-agent systems that can solve complex problems more efficiently than previous technologies.
Arayaman Reddi
Focus on Adversarial Reinforcement Learning
One focus of his work is adversarial reinforcement learning. This involves training systems under difficult, hostile conditions in order to increase their robustness. One example project is robot locomotion, in which robots are trained to move against an opposing agent in simulated adverse environments such as ice or wind.
Another focus is cooperation in multi-agent systems. Here, Reddi is investigating how agents can work together more efficiently through improved communication. In the long term, this approach could lead to systems that can act effectively both as a team and in adverse scenarios.
In search of mathematical foundations
Reddi describes the gap between theory and practice as one of the biggest challenges in current AI research. Many of today’s machine learning methods are based on “fuzzy” ideas without a solid mathematical foundation, says Reddi – this leads to uncertainties in application and troubleshooting, as it can never be proven exactly where the problem lies.
“ChatGPT works well for creating simple texts, but often fails at more complex tasks like writing legal documents or analyzing long texts, such as about medical devices, and it’s hard to figure out why that is,” Reddi says. “There are so many moving parts that we only know work because of empirical evidence, but it’s hard to say they didn’t work because of this or that.”
His aim is therefore also to make a contribution to AI research that is simple and general enough to be implemented – and explained – by other researchers in multi-agent projects.
Multi-agent systems for communication networks
The hessian.AI research network plays a crucial role in Reddi’s work. By collaborating with other researchers from various fields such as computer vision and data management, he has expanded his knowledge and gained new perspectives for his research, Reddi explains in our interview.
When asked, he also talks about the potential social benefits of his research: the use of multi-agent systems in areas such as traffic flow and communication networks in particular could lead to more efficient and environmentally friendly solutions.
He also hopes that his research findings will help other scientists to develop more efficient algorithms and thus have a greater impact on the field of artificial intelligence in general.
The European Association for Computer Graphics has awarded TU Computer Science Professor Justus Thies the prestigious Eurographics Young Researcher Award 2024. The prize is considered one of the most important awards in Europe for computer graphics and is awarded annually to two promising young scientists.
Thies was honored in particular for his groundbreaking work in the field of markerless motion capture and synthesis in computer graphics. His research on facial reenactment and video manipulation has been widely recognized.
This’ research group works at the intersection of computer graphics, computer vision and machine learning. The new AI-based methods focus on markerless motion capture of facial expressions, human bodies and non-rigid objects in general.
Justus Thies has been a full professor for ‘3D Graphics & Vision’ at the Technical University of Darmstadt since 2023 and independently heads the ‘Neural Capture & Synthesis’ research group at the Max Planck Institute for Intelligent Systems in Tübingen. He received his doctorate from the University of Erlangen-Nuremberg in 2017 with a thesis on markerless motion capture of facial representations and its applications. In addition, Justus Thies is part of hessian.AI’s RAI research collective, which is dedicated to the further development of current DL-based AI systems towards “Reasonable AI”.
Photo: Patrick Bal
About Felix Friedrich
Felix Friedrich is a researcher at the Department of Computer Science at the Technical University of Darmstadt. He studied electrical engineering at TU Dortmund University and holds two master’s degrees from TU Darmstadt in Autonomous Systems and Computer Science with a minor in Psychology.
Since 2021, Friedrich has been doing his doctorate at the Machine Learning Lab at the Department of Computer Science at TU Darmstadt.
AI Cognition and machine intelligence
How do humans think and learn? The question of human cognition can be transferred to artificial intelligence and is even a decisive element in the comprehensibility of AI systems. Why does an AI system make a certain decision, for example, and how can the decision-making process be explained so that a human can also understand it?
Friedrich explains that AI systems often use shortcuts and that this results in spurious correlations. He cites the machine recognition of dogs and wolves as an example: If the animal to be recognized is in a forest, the AI system will probably recognize a wolf, but in a domestic context it will recognize a a dog.
AI systems are therefore heavily dependent on human feedback – and the better you understand how the machine works and where the errors lie, the better you understand the potential of AI systems to improve them.
Friedrich emphasizes the pragmatism that characterizes his research: “Today, we have a good idea of what AI does and why. But AI is not perfect. Even though we understand more and more about how it works, the question remains as to how the systems are used. And they need to be improved.”
Friedrich sees great potential for improvement in the fairness and bias of AI models. For example, diffusion models such as image generators often always output the same images for certain age, occupation and gender specifications: The banker is often a white man in a suit, the nurse a woman and the child from Africa in poor circumstances.
“The characteristics of appearance in generated images do not necessarily reflect reality,” says Friedrich. So how do you get more diversity in, and prejudices and beauty ideals out?
Friedrich points to a paradigm: You have to train AI models with data that reflects the whole: “A child who is not supposed to use swear words still needs to know and understand them in order to avoid them.”
AI models are similar, for example with regard to differences in cultural factors. Although targeted prompts can be used to counteract this, they do not always solve the problem of stereotypes or even introduce other stereotypes.
Bias reduction is a balancing act
Friedrich emphasizes a core problem: normative terms such as “right” and “wrong” or “good” and “bad” are difficult to prove in AI models: “It is virtually impossible to train an AI model that has no biases. But you have to look: Where does this cause damage and how can it be prevented?”
A certain amount of influence is possible, says Friedrich: “If you had trained an AI model 20 years ago, a question about whether it is good to fly to New York would probably have been viewed positively. Today, the AI model would probably be more critical of this question because the context has changed. Teach the AI to understand such contexts independently. Create an understanding of normative values.”
Training your own AI models is time-consuming and cost-intensive and only feasible for a small number of companies. For Friedrich, one solution is to overwrite moral values in an AI model, i.e. to adapt them to a personal idea.
In technical jargon, this is called a “revision corpus”: the knowledge of the AI model does not change, it is just queried differently. This depends on the accuracy of the language.
Friedrich again recognizes parallels to humans here, for example in explanations and judgments based on similar cases, examples and situations.
“The challenge is the normative question: What is right and what is wrong? Am I perhaps overwriting other values that I don’t want to overwrite? It is incredibly difficult to remove or reduce biases without introducing new biases.”
Error tolerance and pragmatism in AI development
With “Fair Diffusion”, an adapted diffusion model, Friedrich and his team have developed an image generator that aims to reduce feature bias.
AI models often generate blonde-haired and female people, for example, even though they are underrepresented worldwide. “Fair Diffusion” helps to achieve a fairer distribution of the sexes. At the same time, other attributes such as skin color, hair length or similar can also be influenced.
Removing prejudice from AI models altogether? Probably an impossible undertaking.
These ethical considerations are a central component of Friedrich’s research. The researcher once again appeals to human pragmatism: risks are to be expected with humans, for example in road traffic. It is therefore wrong to expect machines to never make mistakes.
“You have to have a margin for error in any case. But you should still minimize the risk. That is the top priority,” says Friedrich.
For Friedrich, the question of causality or correlation has social implications: “Who actually decides what the world that an AI model reflects should look like?” It is a major challenge when a few large companies own AI models that determine how AI models function and what values they represent.
Friedrich’s research addresses these questions. A recent paper was accepted in the renowned journal “Nature Machine Intelligence”, a great success for the young researcher.
In any case, his work in fairness and biases has met with very positive feedback in the community, such as Fair Diffusion: “My work has a social benefit and high relevance. It gives me a lot of meaning in my everyday life,” says Friedrich.
Friedrich appreciates the research network and infrastructure of hessian.AI | The Hessian Center for Artificial Intelligence : “The infrastructure of hessian.AI | The Hessian Center for Artificial Intelligence is unparalleled in Germany. You can conduct cutting-edge research here, supported by a unique network of researchers and professors.”
The “AI Startup Landscape Hessen 2024” is intended to increase the visibility of AI startups in Hessen and thereby also facilitate access for companies, investors, politicians and those interested in startups. Thanks to the excellent cooperation with other partners from the startup ecosystem in Hessen, the team of AI Startup Rising created a data-based tool that provides a comprehensive overview of AI startups in Hessen.
In contrast to other startup maps, the AI Startup Landscape Hessen also depicts startups in the late pre-foundation phase. This area in particular has a lot of potential in Hesse, says Tobias Kehl, the project leader of AI Startup Rising:
“It is important to us to also make projects visible that have not yet been officially founded. Hesse has a great strength here thanks to its many universities, colleges and support programs. In addition, it is immensely important for these projects in particular to come into contact with potential customers and investors at an early stage.”
The database behind the Landscape is to be further expanded in the longer term to provide more detailed information on the teams, funding, programs completed, etc. In the future, this collected data could help to identify success factors and stumbling blocks and incorporate this knowledge into the funding programs of the ecosystem and authorities.
Take me to the AI Startup Landscape Hessen