Rank3 Code for ISIC 2018: Skin Lesion Analysis Towards Melanoma Detection, Task 3
New!!: Code has been updated with very basic settings. [2019/11/26]
New!!: Code and README would be updated very soon [2019/11/19]
Note: top 2 teams used additional public dataset to train models. However, our code don't use any extra data.
You can submit your prediction to liveboard now, if you want to find some better methods.
Our submission csv and prediction score npy can be downloaded from Google Drive
You can download the dataset from Google Driver.
You can run the code from top directory.
bash ./scripts/xxx.sh
weight_sample_ = np.array([1113,6705,514,327,1099,115,142])/10015
weight_sample_ = 0.05132302/weight_sample_
If you use this code in your research, please cite this project.
@article{zhuang2018skin,
title={Skin lesion analysis towards melanoma detection using deep neural network ensemble},
author={Zhuang, J and Li, W and Manivannan, S and Wang, R and Zhang, JL Jian-Guo and Pan, J and Jiang, G and Yin, Z},
journal={ISIC Challenge 2018},
volume={2},
year={2018}
}