Taylor Arnold, Assistant Professor of Statistics, @statsmaths
Lauren Tilton, Assistant Professor of Digital Humanities, @nolauren
This repository contains notes, code, and data for the DH2019 AVinDH workshop on 8 July 2019 in Utrecht. Feel free to use/share/adopt these notes for other courses.
Our workshop is dedicated to providing a harassment-free experience for everyone. We do not tolerate harassment of participants in any form. If someone makes you or anyone else feel unsafe or unwelcome, please report it as soon as possible. Harassment and other code of conduct violations reduce the value of our event for everyone. We want you to be happy at our event. People like you make our event a better place.
In order for this tutorial to be successful, we ask that participants take note of the following guidelines throughout the session:
- Being open minded including leaving room for exploration and creativity.
- Please try stay at the same point with us in the tutorial. If you are finished with a section ahead of time, you are more than welcome to hack away at our code. We find that staying together through the tutorial works best for everyone involved.
This workshop is broken into two parts. In this first we provide an overview of how images are represented in programming languages and the specific challenges that undergird the computational analysis of visual materials. In the second part, we show how automatic feature extraction can be used to address some of these challenges. We show how the distant viewing toolkit is being developed to assist in this process.
Generally, for workshops that are half-day or longer we tend to require that participants follow along with the code and methods as they are presented. We make sure that everyone is able to run each line of code and debug issues as they arise. When giving an hour-long (or less) workshop, this ideal method is unmanagable and we typically run the workshop as a live demo, giving the code to participants to follow along afterwards.
Today we have an intermediate amount of time (about 1h45) and will trying a hybrid approach. If you would like to setup the code on your laptop and run the code yourself, that great. We will spend some time at the start of the session to get you ready to go. If you are more comfortable just listening, following along with a neighbor, or run into issues that we can't debug in real-time, that is just fine too!
Our schedule for today is:
- Part 1: Basics of Image Process in Python
- Part 2: Deep Learning for Images
- Part 3: Building a Recommendation System
- Part 4: Moving Images and the Distant Viewing Toolkit
For those who would like a deeper dive into image analysis in Python, please check out the notes we have from our week-long workshop at HILT2019.
The materials for today's workshop can be found in the following file:
The code is also available in this GitHub repository, but the data files are too large to store using git.
We will be using the Python programming language for the workshop, as well as several third-party packages. All of it is free and open source. Here is the link to the Anaconda version of Python that we suggest you use:
If you would like to follow along during the workshop, we suggest trying to install this software ahead of the workshop as it can take a bit of time to fully install.
Anaconda Python comes with many of the modules needed for general purpose data science work. For today, you should be able to work with just the default libraries.
You will need to install several others if you want to apply these techniques to your own dataset. The way that you install these is slightly different depending on your platform.
- macOS / linux: open up a terminal window and type
which conda. If should print out a path with "anaconda3" in the name. - windows: anaconda python should have installed a program called Anaconda Prompt. Open this and type the commands below into the prompt.
You can then install the libraries with the following (run each line one at a time):
conda install keras
conda install opencv
conda install tensorflow
pip install keras-retinanet
pip install dvt
Once that is done, you should be able to follow along with most of the tutorial.
A very limited set of materials about image analysis that either directly address humanities applications or are particularly accesible and interesting to a general audience. Please suggest additional references as we would like to expand this list as widely as possible.
- Arnold, Taylor, and Lauren Tilton. Humanities Data in R. New York: Springer, 2015. link.
- Arnold, Taylor, and Lauren Tilton. "Distant viewing: analyzing large visual corpora" Digital Scholarship in the Humanities. link
- Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, Alexander C. Berg and Li Fei-Fei. "ImageNet Large Scale Visual Recognition Challenge". IJCV, 2015. link
- Wevers, Melvin, and Thomas Smits. "The visual digital turn: Using neural networks to study historical images" Digital Scholarship in the Humanities. link