About
Dr Yiming Ma
Focused on multimodal learning for medical imaging AI, with a broader foundation in computer vision.
I am Dr Yiming Ma. My current focus is multimodal learning for medical imaging AI: how models can combine different imaging modalities and complementary sources of information. My broader research background spans computer vision, crowd counting, vision–language models, multimodal driver monitoring, and human-centred visual systems.
I am especially interested in whether multimodal models remain reliable under real constraints: whether they still work when a modality is missing, learn effectively from sparse labels, operate under limited compute, and state the conditions under which an empirical claim holds. This site records the outputs and explains those choices.
“False facts are highly injurious to the progress of science, for they often endure long.”
— Charles Darwin, The Descent of Man · Source for the Darwin quotation
Research interests
- Multimodal medical imaging
- Multimodal representation learning
- Vision–language learning
- Reliable AI under missing modalities
Working principles
- Learn reliably from sparse labels.
- Combine information when sensors fail.
- Consider deployment under limited compute.
- State the conditions and limits of empirical claims.
Contact
For questions about my research or this site, you can contact me by email.