Research Collaboration - HumbleBeeAI x KAIST ICLab

HumbleBeeAI is collaborating with KAIST’s Interactive Computing Laboratory (ICLab) on ongoing research in mobile sensing and human behavior modeling for mental health.

This collaboration explores how data collected from everyday devices can help us better understand long-term behavioral patterns and their relationship to mental well-being. A central focus is working with longitudinal, real-world sensing data collected across different people, contexts, and time periods - the kind of setting where models often struggle due to differences between individuals and changes in daily life.

The research direction includes:

• Building and analyzing multi-user, multi-context mobile sensing data for behavioral and mental health studies
• Improving cross-user generalization (how well models transfer from one person to another)
• Supporting development of models that can be generalizable at the population level while still being personalizable for individuals in real-world applications

Research areas: Mobile sensing, mental health, human behavior modeling, longitudinal datasets, cross-user generalization