DistillPath: An Efficient 22M Distilled Pathology Encoder Approaching Large Foundation Model Performance
DistillPath:一种高效的22M级蒸馏病理编码器,性能接近大型基础模型
机构 * ETH Zürich(苏黎世联邦理工学院)
AI总结 本文提出22M级蒸馏病理编码器DistillPath-KS16,通过蒸馏86M至1.1B的病理编码器得到,性能接近大型模型Virchow2,参数少29倍、速度快25倍以上,在EVA等基准上表现优异。
Comments 26 pages, 5 figures. Accepted at the ECCV 2026 Workshop on Medical Foundation Models and Benchmarks (MedFM-Bench)