Keras 预测 model - 如何获得第一个元素

Question

Hi I am doing a small model which predicts fruits. I have a function which runs a series of predfor various images and they output the prediction as shown below.

[[0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]]

I know the output is not an array, however I would like to know if there is a possible way to check the values at certain positions. For example, if it was an array, i would have done:

if prediction[0] == 1:
    print(prediction, "Apple")

However, since it's not, I have no clue how to check the values inside it. Is there a way I can check?

The function is this:

def fruit_prediction(image_dir):
    img_list = os.listdir(image_dir)
    print(img_list)
    for fruits in img_list:
        path = os.path.join(image_dir, fruits)
        img = image.load_img(path, target_size = (150, 150))
        array = image.img_to_array(img)
        x = np.expand_dims(array, axis=0)

        vimage = np.vstack([x])
        prediction = model.predict(vimage)
        print(prediction, fruits)

Answer 1

Use of np.argmax() may be solve your problem

here np.argmax(prediction ) will return index of highest probability.Now you have index.with the help of index you can easily determine fruit

ex.

#let your food rep index in one hot encoding like below.
fruit={0:'apple',1:'orange',2:'banana'....like that}

Now you have prediction as you calulated in above function

 index=np.argmax(prediction)
 print("fruit name",fruit[index])

Answer 2

Hi Guys I managed to fix this by using prediction.argmax(), like this

if prediction.argmax() == 0:
    print(np.around(prediction, 3), "\n(File: ", fruit, ") Prediction: Apple\n")