Mastering Computer Vision with TensorFlow 2.x

Defining the model architecture

The model involves using a convolutional neural network (CNN) to process the facial image along with its 16 key-points. For details on CNNs, please refer to Chapter 4, Deep Learning on Images. The input to the CNN is train and test images, along with their key-points, and its output will be key-points corresponding to a new image. The CNN will learn to predict the key-points. Details of the model architecture are shown in the following image:

The code for the preceding model is as follows:

model = Sequential()model.add(Conv2D(32, (3, 3), input_shape=(299,299,1), padding='same', activation='relu'))model.add(MaxPooling2D(pool_size=(2, ...