Mastering TensorFlow 1.x

Image classification using retrained  VGG16 in TensorFlow

Now we are going to do the magic of retraining the VGG16 model for the COCO animals dataset. Let us start by defining three placeholders:

• is_training placeholder specifies if we are using the model for training or for predictions
• x_p is a placeholder of shape (None,image_height, image_width,3) for inputs
• y_p is a placeholder of shape (None,1) for outputs

is_training = tf.placeholder(tf.bool,name='is_training')x_p = tf.placeholder(shape=(None,image_height, image_width,3),                    dtype=tf.float32,name='x_p')y_p = tf.placeholder(shape=(None,1),dtype=tf.int32,name='y_p')

As we explained in the strategies section, we shall restore the layers from the checkpoint file except for the last layer, ...