Tag
CAMEL introduces a cross-architecture distillation framework with an asymmetric masking strategy that transfers knowledge from large retinal foundation models like RETFound to ultra-lightweight student models, preserving accuracy while drastically reducing computational cost for resource-limited clinical deployment.
Explores the possibility of seeing new colors by stimulating individual cone cells in the retina, based on a collaboration between computer scientist Ren Ng and vision scientist Austin Roorda.
This paper introduces REVEAL++, a differentiable phenotypic grouping method for vision-language contrastive learning, applied to retinal fundus images and clinical risk narratives for Alzheimer's disease risk prediction, outperforming discrete grouping baselines.