New paper published: Representing and Detecting Label Ambiguity in IMU-Based Exercise Evaluation

We are pleased to announce a new publication from our research group!

In automated movement analysis for physiotherapy — for instance, exercises performed at home without supervision — assessment is not always clear-cut: even trained raters sometimes disagree on borderline cases. Our new paper shows how AI systems can recognize this uncertainty and communicate it transparently, rather than disregarding it. This makes automated assessment not only more accurate, but also more understandable for patients.

📄 Now open access in the journal AI (MDPI):
https://www.mdpi.com/2673-2688/7/9/356

by Andreas Spilz, Heiko Oppel, and Michael Munz

#AI #Physiotherapy #Research #UlmUniversityOfAppliedSciences

Scroll to Top