Pathology segmentation implement with PyTorch.

Network architecture: U-Net, Attention U-Net, R2 U-Net, R2 Attention U-Net, HSU-Net

Segmentation Demo Result:

对比模式

合并模式

---

Dependencies

见各模块的Dependencies

---

文件结构

│  README.MD
│
├─Preprocessing
│  ├─ColorNormalization
|  |
│  |    
│  ├─DataAugmentation
│  │
|  |        
│  └─DataProcess
│          
├─SegmentationViewer
│
│          
└─Training
            

• Preprocessing模块

• SegmentationViewer模块

• Training模块-PyTorch实现

• Training模块-PaddlePaddle实现

---

Thanks

The birth of this project is inseparable from the following projects:

• Pytorch-UNet：PyTorch implementation of the U-Net for image semantic segmentation with high quality images

• PaddleSeg: PaddleSeg是基于PaddlePaddle开发的端到端图像分割开发套件，覆盖了DeepLabv3+, U-Net, ICNet, PSPNet, HRNet, Fast-SCNN等主流分割网络。通过模块化的设计，以配置化方式驱动模型组合，帮助开发者更便捷地完成从训练到部署的全流程图像分割应用。

• VisualDL: VisualDL是飞桨可视化分析工具，以丰富的图表呈现训练参数变化趋势、模型结构、数据样本、高维数据分布等。可帮助用户更清晰直观地理解深度学习模型训练过程及模型结构，进而实现高效的模型优化。

• Fast_WSI_Color_Norm: Codes for Fast GPU-Enabled Color Normalization of Whole Slide Images in Digital Pathology

• ASAP: ASAP is an open source platform for visualizing, annotating and automatically analyzing whole-slide histopathology images.

---