Skip to content
arXiv cs.CV · Papers

OrthKD: Extracting Generalized Clinical Knowledge from Heterogeneous Teachers for Lightweight Deployment

arXiv:2607.25545v1 Announce Type: cross Abstract: Deploying diabetic retinopathy (DR) screening models in primary care requires edge-efficient systems that remain accurate, safe, and reliable under domain shift. Multi-teacher knowledge distillation (KD) is a natural compression strategy, but existing approaches largely