Artificial Intelligence By Example

Running a prediction program

The same CNN_CONCEPT_STRATEGY.py described in Chapter 11, Conceptual Representation Learning, was used. However, some new features were added to make this prediction testing phase comfortable to use and watch. 

The model is loaded and compiled as described in Chapter 11, Conceptual Representation Learning. The identify(target_name) function contains the same philosophy. However, instead of showing the pre-processed image and having to click on it to close the window, a matplotlib plt.close function was added, as in the following code:

 if(display==1):   plt.imshow(original)   plt.show(block=False)   time.sleep(5)   plt.close() 

display is a variable that can be set to 0. In that case, no image will be displayed. However, ...