How Artists can set up their own Neural Network — Part 1 — Installation

Originally published at Jackalope.tech (follow the link to subscribe for more)

This article is meant to help artists, designers, and other non-technical* people set up a neural network on their computer. Here’s an article where I introduce the idea of neural networks and how they can be used by artists.

*I would call this a misnomer but that article isn’t finished yet.

What you will need

1. Windows or Linux PC
2. A powerful graphics card
3. A flash drive with at least 8 gb of space
4. A hard drive with at least 150 gb of free space
5. A wired keyboard and mouse

You will need a Windows or Linux PC. If you have a Mac it is both easier and harder to get this working. It is beyond my scope to explain that though and I’m not really a Mac guy anyway.

You will also need a powerful graphics card. Graphics cards are a type of processor that make the images on your screen. Most people have “integrated graphics cards” which means that the graphics card is built into their computer. You need a “dedicated graphics card” to run a neural network. These graphics cards are used to play video games, run 3D modeling programs, or run VR. I have a GTX 970. You will need something at least this powerful to run the neural network.

My most recent painting. I premiered a first edition print of it at the vurv group show last month.

Installing Ubuntu for Linux

The second thing you need to use a neural network is Ubuntu for Linux. Linux is an operating system, like Windows, Mac, iOS, or Android. It’s the graphical user interface your computer uses to interface with you. It connects your hardware input to the computer’s software. It also connects the computer’s software to the computer’s processing hardware. It’s like a busy office assistant who helps people get what they need in the office. Not all software is compatible with all operating systems. That’s why some applications work on an iPhone but won’t work on an Android. Neural Networks are very cutting edge technology. They aren’t made for consumers. Researchers at Universities make them to learn more about how to make them. Most researchers use Linux because it is free and “open source”. Open Source means that they can modify the code to meet their needs. Windows and Apple will not allow you to do that to their operating systems.

We are going to install Ubuntu. That’s a specific distribution, or flavor, of Linux. It’s relatively user friendly, and it is what the program was made for. To install an operating system you need to partition your drive. That means you separate off a part of your hard-drive space for running Ubuntu. Your operating system currently “formats” or organizes your entire hard-drive. You don’t want two office assistants in the same office because they’ll end up getting all tangled up. We are divvying up the office space so that they can work separately.

• Partition your hard-drive. A tutorial on how to do that can be found here.
• Download the Ubuntu installer here.
• Download the Ubuntu installer here.
• Download the USB Install program here. It will help you run the Ubuntu installer.
• Follow the rest of this tutorial to make a Bootable USB Drive.
• Now follow the rest of this tutorial to actually install Ubuntu. Make sure to install Ubuntu on the drive that you partitioned. DO NOT install Ubuntu on your C Drive. That is the drive that Windows is on and if you overwrite that you will lose all your files.
• Download the Ubuntu installer here.

Once you are done with the installation process you will then find yourself in Ubuntu.

Here’s what my set up looks like. I wanted to make it look cool for this tutorial.

Well that concludes part 1! Next week I’ll be releasing part 2 where I will talk about how to install the neural network Deep Style. Then part 3 I’ll be explaining the process of running the neural network and provide you a free script to help you automate your process and the final part 4 will talk about how to use the neural network material in a mixed media digital painting in Photoshop.

Be sure to subscribe on FB or RSS feed in order to get the latest updates, or subscribe to the newsletter to get that plus extra bonus content!