skip to main contentskip to main menuskip to footer Universität Bielefeld Play Search
skip to main content
skip breadcrumb navigation to main content

Human-centered Artificial Intelligence> Collaborations

CITEC
Foto: CITEC/Universität Bielefeld

Collaborations

Map of the world and Germany with pins on different cities where we have collaborations

Clinical Collaborations

As we strive to translate our research into clinical practice, we collaborate with many clinical partners, both within and outside of academia.

The most prominent international clinical collaborations are with our research partners at Mount Sinai Hospital New York, New York University (NYU) and Nara Institute of Science and Technology (NAIST, Japan). With Prof. Dr. Katharina Schultebrauck's lab from NYU, we are integrating the data donation tool Dona into a study of trauma patients for a better understanding of the impact of social interactions. With Dr. Eugenia Alleva at Mount Sinai, we are currently conducting a study on menstrual pain and resilience, integrating the Digital Stress Test (DST) as a prominent digital measure of resilience, and we will conduct the analysis of the participants' non-verbal behavior under stress. Professor Monica Peruscia Hernandez from the Cybernetics & Reality Engineering lab at NAIST collaborates with us on the investigation of cross-cultural differences in non-verbal behaviors during social interactions.

Our Data Donation tool (Dona) is available in five languages (German, English, Ukrainian, Armenian and Russian). With our partners from the Yerevan State University, Armenia, we plan to collect messaging data from populations who have been exposed to war and displacement.

On a national level, we are working with several partners in the context of the Simulated Interaction Task (SIT). The SIT is a simple and easy-to-use web application to video record participants engaging in a standardized social interaction scenario, allowing for objective social interaction behavior analysis.

In a multicenter study lead by Prof. Dr. Isabek Dziobek and conducted by Simon Kirsch at the Albert-Ludwigs-Universität Freiburg and by Dr. Muyu Lin at the Humboldt-Universität Berlin, we integrated the SIT in a study including patients from the autism spectrum. Applying multimodal and explainable machine learning analysis, we try to find differences in facial expressions, gaze, voice and head movements between these patients and people without such a diagnosis (Paper: On Scalable and Interpretable Autism Detection from Social Interaction Behavior und Paper: Towards the automatic detection of social biomarkers in autism spectrum disorder: introducing the simulated interaction task (SIT)), to eventual support clinicians with objective indicators during the diagnostic process.

Going beyond the characterization of autism spectrum conditions, our partners from Prof. Dr. Isabel Dziobek’s lab at Humboldt-University Berlin and Prof. Dr. Timo Brockmeyer’s lab at Universität Münster are using the SIT in the standard psychotherapeutic diagnostic procedure of their university outpatient clinics. Similarly, officially licensed psychotherapists outside academia like Ralf Jostes support the research with the SIT by integrating it into their standard diagnostic procedure. This allows us to analyze social interactions of patients diagnosed with various conditions like depression, ADHD and social anxiety disorders and to develop algorithms for differentiating social behaviors across different mental conditions.

In parallel, we’ve developed novel versions of the SIT. Together with Prof. Dr. Julia Asbrand and Nadine Vietmeier from Universität Jena and Prof. Dr. Isabel Dziobek and Dr. Simone Kirst from Humboldt-University Berlin. We are currently validating the “Kids-SIT” and aim use it for deriving digital biomarkers of social anxiety disorders in children (Pre-registered study: https://osf.io/tw9ne/).

Together with Prof. Dr. Timo Brockmeyer from Universität Münster, we develop the “ED-SIT” suitable for analyzing social interaction behavior in patients with eating disorders.

Following the idea of transdiagnostic approaches, we work together with Prof. Dr. Michael Rapps research group from Universität Potsdam in the Phenotypic, Diagnostic and Clinical Domain Assessment Network (PDCAN). Using supervised and unsupervised machine learning, we are aiming for computational phenotyping of mental disorders.

If you are a potential clinical partner (clinical psychology department, hospital, psychiatry, ambulance, etc.) and interested in one of our tools or our research in general, you can find more information and collaboration packages on the SIT, Dona and DST pages. You are also welcome to get in touch with us directly by e-mail.

Research Collaborations

Since our research is at the intersection of psychology and computer science, we have fruitful interdisciplinary collaborations with several psychology departments. In the context of our research on stress, we are closely working with Prof. Dr. Oliver Wolf’s lab at Ruhr-Universität Bochum. With his expertise, he supports us in the conceptual development of the Digital Stress Test and the planning of clinical and online studies.
With respect to the collection and analysis of stress related video data, we’re additionally cooperating with the lab from Prof. Dr. Petra Wirtz at Universität Konstanz, Prof. Dr. Gregor Domes at Universität Trier and Ileana Schmalbach and Prof. Dr. Katja Petrowski from Universität Mainz. If you’re also working with stress-related video data or plan to conduct stress induction studies, you can find more information on our stress research and a DST collaboration package here.

Within our discipline of computer science, we also collaborate with different partners on method development:

With Prof. Dr. Davide Mottin at Aarhus University we are currently planning a project on graph networks in the context of social media interaction data. At the University of Bielefeld we are working with Prof. Dr. Barbara Hammer on a machine learning based heartbeat detection project. For the next step of the SIT we will develop a version with an avatar together with Prof. Dr. Stefan Kopp (University of Bielfeld).

Research Centers

We are part of several research centers:

CITEC (Center for Cognitive Interaction Technology), at Universität Bielefeld is a research center dedicated to creating intelligent systems that enable natural human-machine interaction. It brings together experts from diverse fields to develop technologies like interactive robots and AI that understand human behavior.

The TRR 318 is a transregional collaborative research center at Universität Bielefeld and Universitt Paderborn about “Constructing Explainability”. They research   how to make AI decisions more understandable for users. By studying how explanations work, the team aims to design AI systems that allow users to actively engage in understanding AI's decisions, creating more transparent and user-friendly assistance systems. We are part of the network with our research about inclusive Explainable Intelligence.

SAIL (SustAInable Life-cycle of Intelligent Socio-Technical Systems) is an interdisciplinary and interinstitutional collaboration of Universität Bielefeld, Universität Paderborn, Hochschule Bielefeld and Technische Hochschule OWL. SAIL focuses on the full AI life-cycle, emphasizing sustainable, long-term development beyond initial training. The project combines AI research with insights from humanities and social sciences, applying its findings to Industry 4.0 and Intelligent Healthcare. Our focus in the collaboration is improving the method of remote heart rate estimation.

The CoAIJoint Research Center unites researchers from Bielefeld, Bremen, and Paderborn to advance human-centered AI. Combining strengths in cognitive interaction, robotics, and explainable AI, the center focuses on creating AI systems that interact meaningfully with people and support self-aware learning.

Completed Cooperations

We are thankful for inspiration and input by all partners that worked with us in different projects in the past:

back to top