Drawing with Googlenet

Drawing with Googlenet

In my previous post, I showed how can I use deep neural networks to generate image examples of the classes it ' s been tra ined to classify. Since we ' ve already started using deep neural networks on ways they were never intended to being used, let's abuse them some More.

There ' s nothing constraining us to generate image examples of one class at a time. Let's see how happens if we try to generate the class visualizations close to all other, such as for instance a gorilla and a French horn

Gorilla playing an odd-looking french horn

Well, it kind of looks like a gorilla playing the French horn. Or let's try dressing up a gibbon via "mixing" The Gibbon class with some of the clothing classes:

A gibbon in a poncho (left) and an et-looking gibbon in a labcoat (right)

Or What about making some scenic nature drawings, such as some foxes underneath an erupting volcano:

Foxes beneath an erupting volcano

Or a ballpoint pen drawing a castle:

A vague ballpoint pen drawing a castle

These mixes of classes kind of work out, though it should is noted that these is the best selections from a numb Er of mixes I tried. It ' s also tempting to create mixes of animal classes to generate some new kind of monster breeds, but most of the time thi s doesn ' t work so well. Here's some I tried though, a mix of a Scotch terrier and a tarantula, and a mix of a bee and a gibbon:

A slightly spidery looking Scotch Terrier (left) and a slightly gibbon-looking bee (right)

Another fun thing we can does when generatinge images are to does the gradient ascent randomly along paths instead of on a sing Le Point. This is course takes a bit longer time, but it allows us to "draw" with the output, such as for instance drawing a Mountai N Range of Alps:

Or a line of jellyfish:

or a circle of Junco birds:

If we try to fill a larger region with visualizations of a class, we can also apply clipping masks, i.e. forcing s to zero in some pattern during gradient ascent. So we can-instance use letters as clipping masks and try-to-create the alphabet with animals:

An A of Apesa B of Bearsand A C of Cobras

Alright, that's enough abuse of our deep neural network for today. I ' ve just scratched the surface here, but there is several fun ways to use deep neural networks for creative visual work With a bit of experimentation (and lots of patience). I ' m going to put the Ipython notebooks I used to make these examples in the Deepdraw repository as soon as I ' ve cleaned up The code, so stay tuned via Twitter.

Drawing with Googlenet