alsino / creative-applications-ml

All code and materials for the DAAD introductory workshop on machine learning

Creative Applications of Machine Learning –
A Hands-On Workshop for (Absolute) Beginners

This is a repository with all the code and materials for the DAAD Alumni-Workshop at daadgalerie_studio, 21 October 2019, 10am - 4pm.

This workshops draws heavily on the friendly open source projects ml5.js and p5.js, which you should check out.

A special and big shoutout to Yining Shi who developed many of the materials provided here. The curriculum of this workshop was heavily inspired by Yining's introductiory course on Bots and Machine Learning at the School of Machines in the summer of 2019 and her Machine Learning for the Web's class at ITP in New York.

Housekeping

We are collecting all links and non-code materials on our EtherPad. This way we can exchange materials and links on the fly with everybody without having to set up a chat group or slack.

Topics

• What is Machine Learning (ML)?
• What are some of the most important concepts in ML?
• What are common applications of machine learning?
• Why should we engage with creative applications of ML?
• What are tools we can use in creative ML applications?

Agenda

10am - 12pm: Introduction to Machine Learning

• A Quick Introduction to Machine Learning (Presentation)
• Hands-on 1: A very basic introduction to JavaScript and p5.js

12pm - 1pm: Lunch Break

1pm - 4pm: Machine Learning with ml5.js and Runway

• Hands-on 2: Image classification with MobileNet
• Hands-on 3: Pose estimation (PoseNet) with ml5.js
• Hands-on 4: Generate images from text (AttnGan-Runway)
• Next steps: Where to go from here? (Resources)
• Feedback

Get started

To run each code example, open your terminal (HowTo: Mac or Windows) and type in the following commands:

Clone this repository (folder)
$ git clone https://github.com/alsino/creative-applications-ml.git

Go to the folder you have just downloaded
$ cd creative-applications-ml

Go to the specific example folder we are working in, i.e. for the first imageClassification example:
$ cd 1-imageClassification

Start up a simple web server in the current directory (note: you will need to have python installed on your machine)
$ python -m SimpleHTTPServer     # $ python3 -m http.server (if you are using python 3)

Install Python

In case you do not have python installed on your computer, here are the instructions for installation for both Mac and Windows.