RockS2Net is a python project that is a Siamese network for rock classification

pip install tensorflow==2.2.0

pip install numpy==1.19.5

pip install -r requirements.txt

• confusion-matrix==0.1

• Keras==2.3.1

• matplotlib==3.5.2

• numpy==1.19.5

• Pillow==9.3.0

• pip==22.3.1

• scikit-learn==1.0.2

• scipy==1.7.3

• sklearn==0.0.post1

Visit the following link and enter the password to download：

https://pan.baidu.com/s/1xZJ9KzfCGmJ6_nsB8dYiVg

password: 1wj6

1、You can change the datasets and related data here to set different tasks：

train_generator=datagen.flow_from_directory('/home/train_data',target_size=(512,512),batch_size=4, save_format='jpg')

test_generator=datagen.flow_from_directory('/home/test_data',target_size=(512,512),batch_size=4, save_format='jpg')

2、You can add weight files here：

7z densenet121_weights_tf_dim_ordering_tf_kernels_notop.7z -r -o /home

weight_path="/home/.h5"

model.load_weights('\home\.h5')

3、Run python STNdensenet121.py.