Technical Design Space Analysis for Unobtrusive Driver Emotion Assessment Using Multi-Domain Context

CODE alumn Luis Coelho wrote his bachelor’s thesis on using machine learning to recognize driver emotions in cooperation with our partner company Porsche. Luis, along with CODE professor Ulrich von Zadow and a team of researchers from Porsche, then turned this into a paper in one of the best journals in human-computer interaction. What a success story and congratulations for such an outstanding graduation, Luis!

Bethge, D., Coelho, L. F., Kosch, T., Murugaboopathy, S., von Zadow, U., Schmidt, A., & Grosse-Puppendahl, T. (2023). Technical design space analysis for unobtrusive driver emotion assessment using multi-domain context. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 6(4), Article 159, 30 pages. https://doi.org/10.1145/3569466

Programming Education and Learner Motivation in the Age of Generative AI: Student and Educator Perspectives

By CODE professors Samuel Boguslawski, Rowan Deer and Mark Dawson

Boguslawski, S., Deer, R., & Dawson, M. G. (2024). Programming education and learner motivation in the age of generative AI: student and educator perspectives. Information and Learning Sciences. Advance online publication. https://doi.org/10.1108/ILS-10-2023-0163

Evaluating an AI algorithm to detect Bird Sounds for the Dawn Chorus App

Interview by CODE professor Florencia Noriega, who supported the Dawn Chorus App team incorporating an AI algorithm (BirdNet) to automatically detect birds from recordings uploaded by users. Her role involved evaluating the performance of the BirdNet algorithm in detecting birds from the dawn chorus dataset. Evaluating an algorithm’s performance is crucial to assess its effectiveness on such a dataset and to choose appropriate model parameters. In May, the Dawn Chorus app released its first version with AI bird sound recognition. Additionally, CODE students Kinga Janossy and Can Turgay developed an annotation tool for the Dawn Chorus project. This tool is supporting researchers collect quality data to train and improve AI bird detection models.

Forms of Decision Models in Organisations [Manuscript Submitted for Publication]

By CODE professor Florian Grote

The social world of communication is built on decisions. Many of these are microscopic operational decisions, when choosing what to say next in a discussion to keep the conversation going or to use tactics to win an argument. Other decisions might seem big in comparison, when it comes to the direction in which a company is developing its strategy. Such decisions are not taken individually, but they get made in social processes. Organisations are building elaborate decision models, representations of assumed scenarios following from the choices faced in the decision, and are increasingly turning to generative AI in the form of LLMs to make their decisions models more powerful and more dynamic. The paper is an investigation of the basal forms of decision models, their use, and the potential of adding an operational AI perspective to them.