PC-MIL: Decoupling Feature Resolution from Supervision Scale in Whole-Slide Learning
PC-MIL:在全片学习中解耦特征分辨率与监督尺度
机构 * Kahlert School of Computing, University of Utah, Salt Lake City, UT 84112, USA Scientific Computing \& Imaging Institute, University of Utah, Salt Lake City, UT 84112, USA Instapath Inc., Houston, TX 77021, USA Tulane University, Department of Biomedical Engineering, New Orleans, Louisiana, USA
专题命中 病理影像 :pathology(abstract);分类 cs.CV
AI总结 PC-MIL通过解耦特征分辨率与监督尺度,提升全片学习中对解剖结构的建模能力,实现跨上下文的泛化性能提升。
Comments 11 pages, 2 figures, 2 tables. Under review at MICCAI 2026