Art Generation with Neural Style Transfer
Welcome to the second assignment of this week. In this assignment, you will learn about Neural Style Transfer. This algorithm was created by Gatys et al.(2015).
In this assignment, you will:
Implement the neural style transfer algorithm
Generate novel artistic images using your algorithm
Most of the algorithms you've studied optimize a cost function to get a set of parameter values. In Neural Style Transfer, you'll optimize a cost function to get pixel values!
1 - Problem Statement
Neural Style Transfer (NST) is one of the most fun techniques in deep learning. As seen below, it merges two images, namely, a "content" image (C) and a "style" image (S), to create a "generated" image (G). The generated image G combines the "content" of the image C with the "style" of image S.
In this example, you are going to generate an image of the Louvre museum in Paris (content image C), mixed with a painting by Claude Monet, a leader of the impressionist movement (style image S).
2 - Transfer Learning
Neural Style Transfer (NST) uses a previously trained convolutional network, and builds on top of that. The idea of using a network trained on a different task and applying it to a new task is called transfer learning.
Following the original NST paper, we will use the VGG network. Specifically, we'll use VGG-19, a 19-layer version of the VGG network. This model has already been trained on the very large ImageNet database, and thus has learned to recognize a variety of low level features (at the earlier layers) and high level features (at the deeper layers).
Run the following code to load parameters from the VGG model. This may take a few seconds.
The model is stored in a python dictionary where each variable name is the key and the corresponding value is a tensor containing that variable's value. To run an image through this network, you just have to feed the image to the model. In TensorFlow, you can do so using the tf.assign function. In particular, you will use the assign function like this:
This assigns the image as an input to the model. After this, if you want to access the activations of a particular layer, say layer 4_2
when the network is run on this image, you would run a TensorFlow session on the correct tensor conv4_2
, as follows:
3 - Neural Style Transfer
We will build the NST algorithm in three steps:
3.1 - Computing the content cost
In our running example, the content image C will be the picture of the Louvre Museum in Paris. Run the code below to see a picture of the Louvre.
The content image (C) shows the Louvre museum's pyramid surrounded by old Paris buildings, against a sunny sky with a few clouds.
3.1.1 - How do you ensure the generated image G matches the content of the image C?
As we saw in lecture, the earlier (shallower) layers of a ConvNet tend to detect lower-level features such as edges and simple textures, and the later (deeper) layers tend to detect higher-level features such as more complex textures as well as object classes.
We would like the "generated" image G to have similar content as the input image C. Suppose you have chosen some layer's activations to represent the content of an image. In practice, you'll get the most visually pleasing results if you choose a layer in the middle of the network--neither too shallow nor too deep. (After you have finished this exercise, feel free to come back and experiment with using different layers, to see how the results vary.)
Exercise: Compute the "content cost" using TensorFlow.
Instructions: The 3 steps to implement this function are: 1. Retrieve dimensions from a_G:
To retrieve dimensions from a tensor X, use:
X.get_shape().as_list()
2.Unroll a_C and a_G as explained in the picture above
3.Compute the content cost:
What you should remember:
3.2 - Computing the style cost
For our running example, we will use the following style image:
This painting was painted in the style of impressionism.
3.2.1 - Style matrix
In NST, you can compute the Style matrix by multiplying the "unrolled" filter matrix with their transpose:
3.2.2 - Style cost
Exercise: Compute the style cost for a single layer.
Instructions: The 3 steps to implement this function are: 1. Retrieve dimensions from the hidden layer activations a_G:
To retrieve dimensions from a tensor X, use:
X.get_shape().as_list()
2.Unroll the hidden layer activations a_S and a_G into 2D matrices, as explained in the picture above.
3.Compute the Style matrix of the images S and G. (Use the function you had previously written.)
4.Compute the Style cost:
3.2.3 Style Weights
You can combine the style costs for different layers as follows:
We've implemented a compute_style_cost(...) function. It simply calls your compute_layer_style_cost(...)
several times, and weights their results using the values in STYLE_LAYERS
. Read over it to make sure you understand what it's doing.
Loop over (layer_name, coeff) from STYLE_LAYERS:
a. Select the output tensor of the current layer. As an example, to call the tensor from the "conv1_1" layer you would do: out = model["conv1_1"]
b. Get the style of the style image from the current layer by running the session on the tensor "out"
c. Get a tensor representing the style of the generated image from the current layer. It is just "out".
d. Now that you have both styles. Use the function you've implemented above to compute the style_cost for the current layer
e. Add (style_cost x coeff) of the current layer to overall style cost (J_style)
Return J_style, which should now be the sum of the (style_cost x coeff) for each layer.
Note: In the inner-loop of the for-loop above, a_G
is a tensor and hasn't been evaluated yet. It will be evaluated and updated at each iteration when we run the TensorFlow graph in model_nn() below.
How do you choose the coefficients for each layer? The deeper layers capture higher-level concepts, and the features in the deeper layers are less localized in the image relative to each other. So if you want the generated image to softly follow the style image, try choosing larger weights for deeper layers and smaller weights for the first layers. In contrast, if you want the generated image to strongly follow the style image, try choosing smaller weights for deeper layers and larger weights for the first layers
What you should remember:
The style of an image can be represented using the Gram matrix of a hidden layer's activations. However, we get even better results combining this representation from multiple different layers. This is in contrast to the content representation, where usually using just a single hidden layer is sufficient.
3.3 - Defining the total cost to optimize
Finally, let's create a cost function that minimizes both the style and the content cost. The formula is:
Exercise: Implement the total cost function which includes both the content cost and the style cost.
What you should remember:
4 - Solving the optimization problem
Finally, let's put everything together to implement Neural Style Transfer!
Here's what the program will have to do:
Create an Interactive Session
Load the content image
Load the style image
Randomly initialize the image to be generated
Load the VGG16 model
Build the TensorFlow graph:
Run the content image through the VGG16 model and compute the content cost
Run the style image through the VGG16 model and compute the style cost
Compute the total cost
Define the optimizer and the learning rate
Initialize the TensorFlow graph and run it for a large number of iterations, updating the generated image at every step.
Lets go through the individual steps in detail.
Lets start the interactive session.
Let's load, reshape, and normalize our "content" image (the Louvre museum picture):
Let's load, reshape and normalize our "style" image (Claude Monet's painting):
Now, we initialize the "generated" image as a noisy image created from the content_image. By initializing the pixels of the generated image to be mostly noise but still slightly correlated with the content image, this will help the content of the "generated" image more rapidly match the content of the "content" image. (Feel free to look in nst_utils.py
to see the details of generate_noise_image(...)
; to do so, click "File-->Open..." at the upper-left corner of this Jupyter notebook.)
Next, as explained in part (2), let's load the VGG16 model.
To get the program to compute the content cost, we will now assign a_C
and a_G
to be the appropriate hidden layer activations. We will use layer conv4_2
to compute the content cost. The code below does the following:
Assign the content image to be the input to the VGG model.
Set a_C to be the tensor giving the hidden layer activation for layer "conv4_2".
Set a_G to be the tensor giving the hidden layer activation for the same layer.
Compute the content cost using a_C and a_G.
Note: At this point, a_G is a tensor and hasn't been evaluated. It will be evaluated and updated at each iteration when we run the Tensorflow graph in model_nn() below.
Exercise: Now that you have J_content and J_style, compute the total cost J by calling total_cost()
. Use alpha = 10
and beta = 40
.
You'd previously learned how to set up the Adam optimizer in TensorFlow. Lets do that here, using a learning rate of 2.0. See reference
Exercise: Implement the model_nn() function which initializes the variables of the tensorflow graph, assigns the input image (initial generated image) as the input of the VGG16 model and runs the train_step for a large number of steps.
Run the following cell to generate an artistic image. It should take about 3min on CPU for every 20 iterations but you start observing attractive results after ≈140 iterations. Neural Style Transfer is generally trained using GPUs.
You're done! After running this, in the upper bar of the notebook click on "File" and then "Open". Go to the "/output" directory to see all the saved images. Open "generated_image" to see the generated image! :)
We didn't want you to wait too long to see an initial result, and so had set the hyperparameters accordingly. To get the best looking results, running the optimization algorithm longer (and perhaps with a smaller learning rate) might work better. After completing and submitting this assignment, we encourage you to come back and play more with this notebook, and see if you can generate even better looking images.
Here are few other examples:
5 - Test with your own image (Optional/Ungraded)
Finally, you can also rerun the algorithm on your own images!
To do so, go back to part 4 and change the content image and style image with your own pictures. In detail, here's what you should do:
Click on "File -> Open" in the upper tab of the notebook
Go to "/images" and upload your images (requirement: (WIDTH = 300, HEIGHT = 225)), rename them "my_content.png" and "my_style.png" for example.
Change the code in part (3.4) from :
to:
Rerun the cells (you may need to restart the Kernel in the upper tab of the notebook).
You can also tune your hyperparameters:
Which layers are responsible for representing the style? STYLE_LAYERS
How many iterations do you want to run the algorithm? num_iterations
What is the relative weighting between content and style? alpha/beta
6 - Conclusion
Great job on completing this assignment! You are now able to use Neural Style Transfer to generate artistic images. This is also your first time building a model in which the optimization algorithm updates the pixel values rather than the neural network's parameters. Deep learning has many different types of models and this is only one of them!
What you should remember:
Neural Style Transfer is an algorithm that given a content image C and a style image S can generate an artistic image
It uses representations (hidden layer activations) based on a pretrained ConvNet.
The content cost function is computed using one hidden layer's activations.
The style cost function for one layer is computed using the Gram matrix of that layer's activations. The overall style cost function is obtained using several hidden layers.
Optimizing the total cost function results in synthesizing new images.
This was the final programming exercise of this course. Congratulations--you've finished all the programming exercises of this course on Convolutional Networks! We hope to also see you in Course 5, on Sequence models!
References:
The Neural Style Transfer algorithm was due to Gatys et al. (2015). Harish Narayanan and Github user "log0" also have highly readable write-ups from which we drew inspiration. The pre-trained network used in this implementation is a VGG network, which is due to Simonyan and Zisserman (2015). Pre-trained weights were from the work of the MathConvNet team.
Leon A. Gatys, Alexander S. Ecker, Matthias Bethge, (2015). A Neural Algorithm of Artistic Style
Harish Narayanan, Convolutional neural networks for artistic style transfer.
Karen Simonyan and Andrew Zisserman (2015). Very deep convolutional networks for large-scale image recognition
Last updated