3 papers
cs.LG2025
Knockout: A simple way to handle missing inputs
Minh Nguyen, Batuhan K. Karaman, Heejong Kim +3
Deep learning models benefit from rich (e.g., multi-modal) input features. However, multimodal models might be challenging to deploy, because some inputs may be missing at inferenc…
cs.CV2025
RealKeyMorph: Keypoints in Real-world Coordinates for Resolution-agnostic Image Registration
Mina C. Moghadam, Alan Q. Wang, Omer Taub +2
Many real-world settings require registration of a pair of medical images that differ in spatial resolution, which may arise from differences in image acquisition parameters like p…
cs.CV2025
BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration
Alan Q. Wang, Rachit Saluja, Heejong Kim +3
We present a keypoint-based foundation model for general purpose brain MRI registration, based on the recently-proposed KeyMorph framework. Our model, called BrainMorph, serves as…