Machine Learning Engineer Nanodegree

Deep Learning

📑   Practice Project 4: Convolutional Neural Networks

In this notebook, we visualize four activation maps in a CNN layer.

1. Import the Image

%%html
<style>
@import url('https://fonts.googleapis.com/css?family=Orbitron|Roboto');
body {background-color: #add8e6;} 
a {color: darkblue; font-family: 'Roboto';} 
h1 {color: steelblue; font-family: 'Orbitron'; text-shadow: 4px 4px 4px #aaa;} 
h2, h3 {color: #483d8b; font-family: 'Orbitron'; text-shadow: 4px 4px 4px #aaa;}
h4 {color: slategray; font-family: 'Roboto';}
span {text-shadow: 4px 4px 4px #ccc;}
div.output_prompt, div.output_area pre {color: #483d8b;}
div.input_prompt, div.output_subarea {color: darkblue;}      
div.output_stderr pre {background-color: #add8e6;}  
div.output_stderr {background-color: #483d8b;}        
</style>

import cv2
import scipy.misc
import matplotlib.pyplot as plt
%matplotlib inline

# TODO: Feel free to try out your own images here by changing img_path
# to a file path to another image on your computer!
img_path = 'images/udacity_sdc.png'

# load color image 
bgr_img = cv2.imread(img_path)
# convert to grayscale
gray_img = cv2.cvtColor(bgr_img, cv2.COLOR_BGR2GRAY)
# resize to smaller
small_img = scipy.misc.imresize(gray_img, 0.3)

# rescale entries to lie in [0,1]
small_img = small_img.astype("float32")/255

# plot image
plt.imshow(small_img, cmap='gray')
plt.show()

2. Specify the Filters

import numpy as np

# TODO: Feel free to modify the numbers here, to try out another filter!
# Please don't change the size of the array ~ :D
filter_vals = np.array([[-1, -1, 1, 1], [-1, -1, 1, 1], [-1, -1, 1, 1], [-1, -1, 1, 1]])

### do not modify the code below this line ###

# define four filters
filter_1 = filter_vals
filter_2 = -filter_1
filter_3 = filter_1.T
filter_4 = -filter_3
filters = [filter_1, filter_2, filter_3, filter_4]

# visualize all filters
fig = plt.figure(figsize=(10, 5))
for i in range(4):
    ax = fig.add_subplot(1, 4, i+1, xticks=[], yticks=[])
    ax.imshow(filters[i], cmap='gray')
    ax.set_title('Filter %s' % str(i+1))
    width, height = filters[i].shape
    for x in range(width):
        for y in range(height):
            ax.annotate(str(filters[i][x][y]), xy=(y,x),
                        horizontalalignment='center',
                        verticalalignment='center',
                        color='white' if filters[i][x][y]<0 else 'black')

3. Visualize the Activation Maps for Each Filter

from keras.models import Sequential
from keras.layers.convolutional import Convolution2D
import matplotlib.cm as cm

# plot image
plt.imshow(small_img, cmap='gray')

# define a neural network with a single convolutional layer with one filter
model = Sequential()
model.add(Convolution2D(1, (4, 4), activation='relu', input_shape=(small_img.shape[0], small_img.shape[1], 1)))

# apply convolutional filter and return output
def apply_filter(img, index, filter_list, ax):
    # set the weights of the filter in the convolutional layer to filter_list[i]
    model.layers[0].set_weights([np.reshape(filter_list[i], (4,4,1,1)), np.array([0])])
    # plot the corresponding activation map
    ax.imshow(np.squeeze(model.predict(np.reshape(img, (1, img.shape[0], img.shape[1], 1)))), cmap='gray')

# visualize all filters
fig = plt.figure(figsize=(12, 6))
fig.subplots_adjust(left=0, right=1.5, bottom=0.8, top=1, hspace=0.05, wspace=0.05)
for i in range(4):
    ax = fig.add_subplot(1, 4, i+1, xticks=[], yticks=[])
    ax.imshow(filters[i], cmap='gray')
    ax.set_title('Filter %s' % str(i+1))

# visualize all activation maps
fig = plt.figure(figsize=(20, 20))
for i in range(4):
    ax = fig.add_subplot(1, 4, i+1, xticks=[], yticks=[])
    apply_filter(small_img, i, filters, ax)
    ax.set_title('Activation Map for Filter %s' % str(i+1))

Using TensorFlow backend.