Here is a combined short summary on my travels to the city of Prague in the Czech Republic along with corresponding images created using Google’s DeepDreams.
What is this DeepDreams you speak of?
Basically, DeepDream is a deep neural network that was trained to recognize objects from millions of images. A deep neural network is composed of a stack of layers. Basically, these layers learn image filters that when applied to an image classify the image (e.g., is this an image of a cat or a dog?).
You give DeepDream an image and specify a layer in the neural network. The original image is then slightly perturbed to create a modified image that causes the specified layer in the neural network to be more activated.
Early layers in the neural network are sensitive to low level concepts like the edges and textures in the image. So if you specify an early layer, your image will be modified to have edges and textures that most activate the early selected layer.
Later (or deeper) layers in the neural network are activated when they see higher level concepts such as faces. So any areas in the original image that slightly look like a face, will be modified to look more like a face.
Okay, but now you might ask, but what about Prague? How was your trip? Did you like the city?
Yeah it was nice! Thanks for asking. Did you want to see some pictures? Here’s one of an old building.
Let’s try some deep dreaming on this. We’ll use the neural network known as VGG16 (it’s a famous neural network that performs very well on competitions). We’ll start by telling VGG16 (the neural network) to modify this image so that one of it’s middle layers becomes more activated. Specifically, we will activate layer conv3_1
from VGG16 (if you don’t know what conv3_1 means, that’s okay – it’s just a technical detail specifying what layer to use). This gives us this:
Now if we activate a deeper layer, conv5_2
, we get this crazy looking image,
Note how the above image actually starts creating a new scene. It changed an ally on the right side to be another door. The windows are all of a different architecture. It started adding plants at the top floor of the main building. The top of the building looks more like a ship now. It’s starting to hallucinate new buildings on the skyline. And the sky! What is happening in the sky?!
Okay enough of this picture. Do you want to see one of the famous Prague clock?
Our tour guide was very excited to show us this clock.
Selecting layer conv4_2
, we see that this layer responds well to circular things…
While going into the deeper layers conv5_1, we capture some disturbing looking animals,
We can create some crazy looking images again using conv5_2
,
How about the inside of a cathedral? You ever seen one before? I hadn’t.
Deep dreaming it…
Strangely, an omnipotent looking reptilian eye comes out of the cathedral’s ceiling using layer conv5_2
,
Do you have any pictures of you in front of Prague castle? You bet I do!
Not sure how to describe what this layer is activated by…
Finally, here’s something you might see if you looked in a mirror while you are dreaming,
I’m struck that the rest of the image is not very distorted. The distortion here is focused mainly on my face.
That’s enough travel pictures and dreaming for me today. Oh! And if you were wondering why I was in Prague, you can read about that here.