-
Folder "CNN_code_Main" This folder includes the major code of the four CNN models for the four tasks.
-
Folder "CNN_graph": this folder includes the code to generate the accuracy-loss VS number of epochs curves of the four models.
-
Folder "RF_model-main": this folder includes the code of the random forest models for the four tasks.
-
Folder "RF_Graph": this folder includes the code to generate the accuracy-max_tree_depth curves for the four models, and a folder named "Graph files" that includes the four graphs of curves.
-
Folder "Tree structures": this folder contains the graphs of the tree structures of the random forest models for the four tasks. Tree parameters are consistent to the code in folder "RF_model-main".
-
Folders "Before the model" and "Random Forest" include some of the draft code - there are some bugs. Please refer to the code in "CNN_code_Main" and "RF_model-main".
-
I am sorry that the model files of the CNN models are too large to be allowed to upload, so these files do not show here. If you require them, please contact me at [email protected].
Requirements to run the code: opencv==3.4.2 scikit-learn==0.24.2 numpy==1.19.2 tensorflow-gpu==1.8.0 keras==2.1.6 matplotlib==3.3.4