Understanding Generalizability of a Wearable System for Automated Detection of Social Interactions in Real-World Environments

Social interactions can improve mental health, while negative interactions can have a detrimental impact. Depending on mental health, the interaction patterns can vary. Therefore, understanding real-world interactions can be a potential for just-in-time adaptive interventions to improve mental health. In this project, we will explore our developed on-device interaction detection system for a deeper understanding of generalizability in Bangladesh, which may play a key role in bringing an impactful system to a country residing LMICs.  Besides, our system can support researchers in disciplines such as behavioral science and psychology to incorporate such a system while exploring their own research problems.