Artificial Intelligence By Example

Training and testing the model

The model was trained by using the same 01_CNN_STRATEGY_MODEL.py program as in Chapter 9, Getting Your Neurons to Work.

Just be sure the directory in the header of the following program is pointing to /dataset, which is scenario number 2:

A=['dataset_O/','dataset_traffic/','dataset/'] scenario=2 #reference to Adirectory=A[scenario] #transfer learning parameter (choice of images)print("directory",directory)

The model is stored in /dataset/model. To test the model, the CNN_STRATEGY_MODEL.py improved in the previous chapter was used. Just change the messages and limit the frame classification loop to 2 as shown in the following code snippet:

MS1='full'MS2='space'I=['1','2','3','4','5','6']for im in range(2):