Michael Wan

I'm a Research Scientist at the Institute for Experiential AI at Northeastern University. I work out of Northeastern's Roux Institute campus in Portland, Maine.

My research is in machine learning and health. I develop deep learning and computer vision tools to advance health research, particularly in developmental health and biomedical imaging, in close collaboration with researchers in psychology and medicine.

I grew up in Toronto, did my HonBSc in math, physics, and philosophy at the University of Toronto, and completed my PhD in mathematical logic at the University of California, Berkeley.

mi.wan@northeastern.edu | Google Scholar | LinkedIn

Research

Computer Vision for Human Health & Development

I work on human pose and state estimation, particularly in adapting algorithms trained on large amounts of mostly adult subject data to the data-scarce infant domain. This includes work on estimating infant faces (ICPR'22), 3D body poses (CVM 2025), postures (ICPRW'22), actions (CVPRW'23), respiration (MICCAIW'23), and sleep (WACVW '25), with a mix of techniques in representation learning, weak supervision, and spatiotemporal neural networks.

I have also worked with developmental psychologists and physical therapists on applications of these techniques to measure infant pacifier sucking (MICCAI'23, CVIU 2024) as a biomarker of developmental delays; postural symmetry (WACV'23) towards early detection of autism spectrum disorder; upper body symmetry (FGW'23) as a sign of torticollis conditions; and infant crib safety (FG'24).

Collaborators: Sarah Ostadabbas, Briana Taylor, Marie Hayes, Emily Zimmerman, Bethany Tunik

Microscopy Image Analysis

More recently, I have been working with cellular and genetic biologists on applying computer vision and medical imaging tools to analyze fine-grained cellular morphology (CVPRW'24), towards understanding cell functions and phenotypes and developing treatments for diseases like osteoporosis.

Collaborators: Christine Lary, Katherine Motyl, Candice Byers

Causality, Domain Adaptation, Explainability, Fairness, Privacy

Even more recently, I have been thinking about causal inference and causal discovery in machine learning, and how these frameworks might be able to help machine learning efforts in domain adaptation, explainability, fairness, and privacy. Stay tuned, or better yet, help me think through these questions!

Collaborators: Zulqarnain Kahn, Resmi Ramachandranpillai

Mathematical Logic

Before getting into machine learning, I did research in mathematical logic. In my PhD thesis with Tom Scanlon, I explored connections between model theory, a branch of metamathematics concerned with formal logical theories and structures instantiating them, and complex geometry, which plays a foundational role in both mathematics and theoretical physics.

Lab Members

Current

Harshil Bhojwani, MS Computer Science

John Gill, MS Bioinformatics

Hardik Bishnoi, MS Computer Science

Liyang Song, MS Data Science

Dilshath Shaik, MS Data Science, 2024

Past

Briggs Twitchell, MS Computer Science, 2024

Kunal Wagle, BS Computer Science

Meghana Chillara, MS Data Science (CVPRW'24)

Brendan Martin, MS Computer Science, 2024 (CVPRW'24)

Sai Kumar Reddy Manne, MS Computer Science, 2024 (MICCAIW'23, CVPRW'24, MS Thesis, CVIU (2024), Preprint, WACVW'25)

Kashish Jain, MS Data Science, 2022 (MS Thesis, MICCAI'23)

Samuel Zlota, BS Computer Science, 2022 (MICCAI'23)

Future

I am looking for students with experience in machine learning, deep learning, computer vision, mathematics, or statistics, and who are interested in working on projects related to my research. Students will be aiming to advance research in the field and publish their results. I make an effort to (i) connect students with projects that match their holistic interests and goals, (ii) provide students with guidance and support needed for them to take their research up to the next level, and (iii) respect students' time and school-work-life balance. If you are interested in working with me, please send me an email with your CV and a short description of your background and interests.

Awards & Funding

2024 | Northeastern University TIER 1 Seed Grant, "AI-Enhanced Vision-Based Privacy-Preserving Infant Sleep Safety Monitoring," with Briana Taylor, Sarah Ostadabbas, Xiaolin Xu.

2023 | Oral Presentation, 26th International Conference on Medical Image Computing and Computer-Assisted Intervention, Workshop in Perinatal, Preterm and Pædiatric Image Analysis (MICCAI PIPPI'23) for S.K.R., Zhu, S., Ostadabbas, S., Wan, M., "Automatic Infant Respiration Estimation from Video: A Deep Flow-based Algorithm and a Novel Public Benchmark."

2022 | Best Scientific Paper Award, 26th International Conference on Pattern Recognition (ICPR'22), for Wan, M. et al., "InfAnFace: Bridging the Infant–Adult Domain Gap in Facial Landmark Estimation in the Wild."

2022 | Northeastern University Spark Fund, "NeuroSense Diagnostics: A smartphone application employing groundbreaking technology for remote infant sucking pattern measurement," VP Software & Operations, with Emily Zimmerman, Sarah Ostadabbas.

2022 | Mathworks Microgrant, "Forecasting Infant Sleep States from Nonnutritive Sucking Patterns."