HoVer-NeXt: A Fast Nuclei Segmentation and Classification Pipeline for Next Generation Histopathology - Model Weights

This repository contains trained weights for: HoVer-NeXt: A Fast Nuclei Segmentation and Classification Pipeline for Next Generation Histopathology (Under review at MIDL2024) More information and code are available at https://github.com/digitalpathologybern/hover_next_inference lizard_convnextv2_large,lizard_convnextv2_base and lizard_convnextv2_tiny correspond to HoVer-NeXt_Large,HoVer-NeXt_Base and HoVer-NeXt_Tiny. pannuke_convnextv2_tiny_* corresponds to the weights trained on the specific pannuke split.本仓库收录了以下模型的训练所得权重： HoVer-NeXt：面向下一代组织病理学的快速细胞核分割与分类流水线 （已提交至MIDL 2024会议并处于审稿阶段） 更多相关信息与代码可访问：https://github.com/digitalpathologybern/hover_next_inference lizard_convnextv2_large（对应ConvNeXt V2架构）、lizard_convnextv2_base 及 lizard_convnextv2_tiny 分别对应 HoVer-NeXt_Large、HoVer-NeXt_Base 与 HoVer-NeXt_Tiny。 pannuke_convnextv2_tiny_* 对应于在指定PanNuke数据集（pannuke）划分上训练所得的模型权重。