use.py

import tensorflow as tf
import keras_preprocessing
from keras_preprocessing import image
import numpy as np

def list_file_of_dir(dir):
    import os
    list_files = os.listdir(dir)
    for file in list_files:
        print(file)

model = tf.keras.models.load_model('rps.h5')

if __name__ == "__main__":
    while(1):
        usr_dir_choose = input("Please input the directory you want to list(scissors/rock/paper): ")
        list_file_of_dir("./rps-test-set/" + usr_dir_choose)

        usr_file_choose = input("Please input the file you want to predict (paste his name here):")
        path = "./rps-test-set/" + usr_dir_choose + "/" + usr_file_choose
        print(path)
        img = image.load_img(path, target_size=(150, 150))
        x = image.img_to_array(img)
        x = np.expand_dims(x, axis=0)
        images = np.vstack([x])
        classes = model.predict(images, batch_size=10)
        result = classes.argmax()
        print(path)
        if result == 0:
            print("it's a paper!")
        if result == 1:
            print("it's a rock!")
        if result == 2:
            print("it's a scissor!")
        print("Do you want to continue? (y/n)")
        usr_continue = input()
        if usr_continue == "n":
            break
        else:
            continue