Custom-CNN-with-EfficientNetV2B0-and-Vision-Transform

Breast Cancer Histopathological Database (BreakHis)

Custom CNN with EfficientNetV2B0 and Vision Transform

The official code for "BreakHis".

Updates

Citation

@article{
  year={2023}
}

How to use

first download models and save them in same directory with IPYNB file as jupyter notebooks then Run nootbooks.

Model weights

Kaggle Drive Link:().

Training and Testing

1. Download the face Brats-2021 dataset from here.

2. 2. Run the following code to install the Requirements.

   pip install -r requirements.txt

3. Run the below code to train the Unet++ with ELU activision as Decoder and... as Encoder with this dataset.

breast-cancer-histopathological-breakhis (2)

6. Test trained model with this dataset in in IPYNB too.

Results

Performance comparision on Classification dataset.