This repository contains an implementation of the following paper: :point_down:

V. K. Singh et al., "FCA-Net: Adversarial Learning for Skin Lesion Segmentation Based on Multi-Scale Features and Factorized Channel Attention," in IEEE Access, vol. 7, pp. 130552-130565, 2019, doi: 10.1109/ACCESS.2019.2940418.

FCA-Net

Prerequisites

• Linux
• Python with numpy
• NVIDIA GPU + CUDA 8.0 + CuDNNv5.1
• pytorch 4.0/4.1
• torchvision

Getting Started

+ Clone this repo:

cd Skin-Project

+ Get dataset

unzip dataset/skin.zip

+ Train the model:

python train.py --dataset skin --nEpochs 200 --cuda

+ Test the model:

python test.py --dataset skin --model checkpoint/skin/netG_model_epoch_100.pth --cuda

Citation:

If you use the code in your work, please use the following citation:

@ARTICLE{8832175,
  author={V. K. {Singh} and M. {Abdel-Nasser} and H. A. {Rashwan} and F. {Akram} and N. {Pandey} and A. {Lalande} and B. {Presles} and S. {Romani} and D. {Puig}},
  journal={IEEE Access}, 
  title={FCA-Net: Adversarial Learning for Skin Lesion Segmentation Based on Multi-Scale Features and Factorized Channel Attention}, 
  year={2019},
  volume={7},
  number={},
  pages={130552-130565},
  doi={10.1109/ACCESS.2019.2940418}}