Referring to Convolutional Neural Network (CNN) code in below cnn_model = models.Sequential() cnn_model.add(layers. Conv2D(16, (11,11), activation='relu',...

 

Transcribed Image Text:

Referring to Convolutional Neural Network (CNN) code in below cnn_model = models.Sequential() cnn_model.add(layers. Conv2D(16, (11,11), activation='relu', input_shape=(256,256,3))) 2))) cnn_model.add(layers.MaxPooling2D((2, cnn_model.add(layers. Conv2D(32, (7,7), activation='relu')) 2))) cnn_model.add(layers. Conv2D(64, (1,1), activation='relu')) cnn_model.add(layers. Conv2D(128, (5,5), activation='relu')) cnn_model.add(layers.MaxPooling2D((2, 2))) cnn_model.add(layers.MaxPooling2D((2, cnn_model.add(layers.Conv2D(256, (3, 3), activation='relu')) cnn_model.add(layers.MaxPooling2D((2, 2))) cnn_model.add(layers.Conv2D(512, (3, 3), activation='relu')) cnn_model.add(layers.Flatten()) cnn_model.add(layers.Dense(512, activation='relu')) cnn_model.add(layers. Dense(1, activation='sigmoid')) a. Illustrate the model structure with details of layer labeling. (3 marks) b. Analyze image output shape and its total trainable parameters for each stack of the CNN layer. (20 marks) c. Illustrate the new model structure if VGG pre-trained model is to be integrated to the existing CNN structure. (2 marks) Referring to Convolutional Neural Network (CNN) code in below cnn_model = models.Sequential() cnn_model.add(layers. Conv2D(16, (11,11), activation='relu', input_shape=(256,256,3))) 2))) cnn_model.add(layers.MaxPooling2D((2, cnn_model.add(layers. Conv2D(32, (7,7), activation='relu')) 2))) cnn_model.add(layers. Conv2D(64, (1,1), activation='relu')) cnn_model.add(layers. Conv2D(128, (5,5), activation='relu')) cnn_model.add(layers.MaxPooling2D((2, 2))) cnn_model.add(layers.MaxPooling2D((2, cnn_model.add(layers.Conv2D(256, (3, 3), activation='relu')) cnn_model.add(layers.MaxPooling2D((2, 2))) cnn_model.add(layers.Conv2D(512, (3, 3), activation='relu')) cnn_model.add(layers.Flatten()) cnn_model.add(layers.Dense(512, activation='relu')) cnn_model.add(layers. Dense(1, activation='sigmoid')) a. Illustrate the model structure with details of layer labeling. (3 marks) b. Analyze image output shape and its total trainable parameters for each stack of the CNN layer. (20 marks) c. Illustrate the new model structure if VGG pre-trained model is to be integrated to the existing CNN structure. (2 marks)

Answer rating: 100% (QA)

The answer provided below has been developed in a clear step by step mannerStep 1 Solution a Illustrate the model structure with details of layer labeling 1Layer 1 Conv2D layer with 16 filters of size ... View the full answer