Multimodal Machine Learning for supporting attachment-based intervention
Collaborating with Dr. Lisa Berlin‘s team at the University of Baltimore School of Socia Work, we are exploring the feasibility of automating the process of characterizing the parent-child interaction using video data collected at home from the Attachment-based intervention. We are experimenting with a suite of multimodal machine learning analysis methods. This project lays the foundation for future research in developing data-driven and human-centered tools to support social workers in delivering home-based interventions.
Collaborator: Dr. Lisa Berlin and her teams (University of Maryland School of Social Work)
Student Researchers:
- Katherine Guerrerio (Johns Hopkins CS undergraduate, Fall 2023 – current)
- Atefeh Jebeli (UMBC IS Master student, Spring 2022 – Summer 2023)
- Sophia Papparotto (UMBC CS Undergraduate student, Spring 2022)
Publications:
- Guerrerio, K.M*., Chen, L.K., Berlin, L.J., & Harden, B.J. Causal Explanation of the Quality of Parent-Child Inter- actions with Multimodal Behavioral Features. In Proceedings of the 27th International Conference on Multimodal Interaction (ICMI ’25), October 13– 17, 2025, Canberra, ACT, Australia. ACM, New York, NY, USA, 11 pages. https://doi.org/10.1145/3716553.3750816
- Jebeli, A*., Chen, L.K., Guerrerio, K *., Papparotto, S*., Berlin, L ^., & Jones Harden, B ^. (2024). Quantifying the Quality of Parent-Child Interaction Through Machine-Learning Based Audio and Video Analysis: Towards a Vision of AI-assisted Coaching Support for Social Workers. ACM Journal of Computing and Sustainable Societies, 2(1), Article 6, 21 pages.