behavioral_cloning.py 文件源码

def SmallNetwork(input_shape):
    model = Sequential()
    # 2 CNNs blocks comprised of 32 filters of size 3x3.
    model.add(ZeroPadding2D((1, 1), input_shape=(img_width, img_height, 3)))
    model.add(Convolution2D(32, 3, 3, activation='elu'))
    model.add(ZeroPadding2D((1, 1)))
    model.add(Convolution2D(32, 3, 3, activation='elu'))
    # Maxpooling
    model.add(MaxPooling2D((2, 2), strides=(2, 2)))

    # 2 CNNs blocks comprised of 64 filters of size 3x3.
    model.add(ZeroPadding2D((1, 1), input_shape=(img_width, img_height, 3)))
    model.add(Convolution2D(64, 3, 3, activation='elu'))
    model.add(ZeroPadding2D((1, 1)))
    model.add(Convolution2D(64, 3, 3, activation='elu'))
    # Maxpooling + Dropout to avoid overfitting
    model.add(MaxPooling2D((2, 2), strides=(2, 2)))
    model.add(Dropout(0.5))    

    # 2 CNNs blocks comprised of 128 filters of size 3x3.
    model.add(ZeroPadding2D((1, 1), input_shape=(img_width, img_height, 3)))
    model.add(Convolution2D(128, 3, 3, activation='elu'))
    model.add(ZeroPadding2D((1, 1)))
    model.add(Convolution2D(128, 3, 3, activation='elu'))
    # Last Maxpooling. We went from an image (64, 64, 3), to an array of shape (8, 8, 128)
    model.add(MaxPooling2D((2, 2), strides=(2, 2)))    

    # Fully connected layers part.
    model.add(Flatten(input_shape=input_shape))
    model.add(Dense(256, activation='elu'))
    # Dropout here to avoid overfitting
    model.add(Dropout(0.5))    
    model.add(Dense(64, activation='elu'))
    # Last Dropout to avoid overfitting
    model.add(Dropout(0.5))
    model.add(Dense(16, activation='elu'))    
    model.add(Dense(1))

    return model