从加载的张量流分类模型中获取预测，.pd 文件

Question

Apologies if this is a trivial question. I just can't quite figure it out. I followed the tutorial found here: https://www.tensorflow.org/tutorials/images/classification . Built the model and trained it. I saved the model to disk with: <model_name>.save(<directory location>)

This saved the following files in 'directory location':

A folder 'assets', this folder is empty. A folder 'variables' and 'saved_model.pb'

Using:

loaded_model = tf.keras.models.load_model('directory location', compile=True)

I load the saved model. Then with

loaded_model.summary()

I get:

_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d (Conv2D)              (None, 150, 150, 16)      448       
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 75, 75, 16)        0         
_________________________________________________________________
dropout (Dropout)            (None, 75, 75, 16)        0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 75, 75, 32)        4640      
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 37, 37, 32)        0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 37, 37, 64)        18496     
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 18, 18, 64)        0         
_________________________________________________________________
dropout_1 (Dropout)          (None, 18, 18, 64)        0         
_________________________________________________________________
flatten (Flatten)            (None, 20736)             0         
_________________________________________________________________
dense (Dense)                (None, 512)               10617344  
_________________________________________________________________
dense_1 (Dense)              (None, 1)                 513       
=================================================================
Total params: 10,641,441
Trainable params: 10,641,441
Non-trainable params: 0 

This makes me think the model loaded correctly. Here I get stuck. How do I feed this model a single image and get a prediction, in this case cat or dog? I have searched the web and various answers here and I got as far as:

prediction = loaded_model.predict(???)

I do not know what the ??? needs to be. A concrete example of how to get a prediction for an image file would be greatly appreciated.

Answer 1

According to the following example the input shape for the model is 150,150,3 so in order to make a prediction we need read the image and resize it to make a prediction. (Assuming your image is in RGB or BGR)

import cv2

preprocessed_image = cv2.imread("Path to Image")
preprocessed_image = cv2.resize(preprocessed_image,(150,150))
preprocessed_image = preprocessed_image.reshape(1,150,150,3)
prediction = loaded_model.predict(preprocessed_image)