Award as “AI Newcomer 2023”

AI has the potential to make our everyday lives easier. But how can we succeed in making interaction with future AI as simple as possible and also give older and less tech-savvy people intuitive access to AI systems? Susanne Trick is doing her doctorate on this question at the Centre for Cognitive Science at TU Darmstadt.

“My goal is to enable the most intuitive and natural interaction possible between people and AI systems such as assistance robots. To do this, I use multimodal data such as gestures, speech or gaze directions so that the AI understands the human as best as possible,” says the scientist, describing the focus of her work.

Trick's interdisciplinary research at the interface between cognitive science and AI contributes to the creation of responsible AI systems that can support people without barriers to use. While we humans process different sensory data quite intuitively, our brains must also compute such information algorithmically. In human perception, uncertain information flows less into this calculation than certain information, since the latter is also less reliable. Very similarly, Susanne Trick has developed probabilistic models for the optimal combination of classifiers, which can be used not only for human-AI interaction, but also for a variety of other AI applications.

Her research results have been published at renowned international AI and robotics conferences and in professional journals. In practice, the methods she developed have already been applied in a pilot study on the use of assistance robots in a retirement home in the BMBF project „Kobo34“ and are currently being used for the optimal fusion of classifiers and expert opinions in AI-supported decision-making processes in the BMBF project „IKIDA“. The head of the IKIDA team is Dr.-Ing. Dorothea Koert, who was herself awarded the title of AI Newcomer in 2019.

Ten newcomers were chosen partly by a public online vote and partly by a jury of experts, including former winners of the AI Camp. Part of the jury this year was Michael Lutter, who, like Koert, received the Newcomer*in Award in 2019 and was also doing research at TU Darmstadt at the time. The award helped him, above all, to present his work to several European research groups. This is also his advice to this year's winners: “I would recommend to the AI newcomers to use the network and thereby get more reach for their own research. We should all continue to support each other so that the worldwide attention for 'AI made in Germany' increases even more,” says Lutter, who now conducts research at Boston Dynamics.

The honour makes the work of the up-and-coming talents visible, and thus the relevance and opportunities of AI in a wide variety of fields. The award also gives the newcomers access to an international network of researchers and renowned AI experts from all disciplines.