Program

9:00 – 9:15 Introductions

9:15 – 10:10 Invited talk: Stefan Lüdtke, University of Rostock: Thriving on Less: Coping with Label Scarcity in Sensor-based Human Activity Recognition

10:10 – 10:30 Best paper (Towards Practical, Best Practice Video Annotation to Support Human Activity Recognition)

10:30 – 11:00 – Coffee break

11:00 – 11:45 Session 1: Activity recognition

Retrieval-based Annotation for Multi-Channel Time Series Data of Human Activities

Towards Standardized Dataset Creation for Human Activity Recognition: Framework, Taxonomy, Checklist, and Best Practices

11:45 – 12:30 Session 2: Annotation for healthcare applications

Relation Extraction from Real-World Unstructured Text in the Domain of Dementia

Assessing privacy-friendly local open-source voice annotation for participants with Parkinson’s disease

12:30 – 14:00 Lunch

14:00 -15:10 Session 3: Ergonomics and data synthesis

Automating Ergonomics: Scalable AI for Technical Hand Grip Classification

Recommendations for Datasets Creation Process for Human Motion Generation in Industrial Simulations

Toppled Realities: Challenges in Generation and Validation of Synthetic Data

15:10 – 15:30 Session 4: Annotation of textual data

Large Language Models Rival Human Performance in Historical Labeling

15:30 – 16:00 Coffee break

16:00 – 16:45 Session 4: Annotation of textual data (continued)

Annotation and label validation of upper-tier tribunal decisions in immigration law

Span-Level Domain-Specific Annotated Student Feedback Pilot Dataset

16:45 – 17:15 Discussion: Future Directions and Trends in Annotation for AI Systems

17:15 – 17:30 Closing remarks