Awesome

Duplex Contextual Relations for Polyp Segmentation

Introduction

The repository contains the PyTorch implementation of "Duplex Contextual Relations for PolypSegmentation"

Overview

1. Framework

<img src="imgs/framework.png"/> Figure 1: Overview of our proposed DCRNet.

2. Quantitative Results

<img src="imgs/visualization.png"/> Figure 2: Qualitative results.

3. Qualitative Results

<img src="imgs/EndoScene.png"/> Figure 3: Quantitative results on EndoScene dataset. <img src="imgs/Kvasir-SEG.png"/> Figure 4: Quantitative results on Kvasir-SEG dataset. <img src="imgs/PICCOLO.png"/> Figure 5: Quantitative results on PICCOLO dataset.

Usage

1. Prerequisite environment

torch>=1.5.0
torchvision>=0.6.0
tqdm
scipy
scikit-image
PIL
numpy
CUDA

2. Dataset downloading

Downloading the CVC-EndoSceneStill dataset, which can be found in this Google Drive link
Downloading the Kvasir-SEG dataset, which can be found in this Google Drive link
To access the PICCOLO dataset, please visit here

3. Train

Assign your customized path of --train_path ,--save_root and --gpu in Train.py.
Run python Train.py

4. Test

Assign the --pth_path , --data_root , --save_root and --gpu in Test.py.
Run python Test.py
The quantitative results will be displayed in your screen, and the qualitative results will be saved in your customized path.

5. Evaluate

The evaluation code is stored in ./utils/eval.py
You can replace it with your customized evaluation metrics.