Text-to-image generation

Dear friends,

At the Eep Talstra Centre for Bible and Computer (www.etcbc.nl) Hidde Makimei, a student from computer science, is doing a project on AI generated images from biblical texts. Whereas some of us may just use DALL-E or Midjourney for fun, and others stand in awe when looking at https://www.openbible.info/labs/ai-bible-art/, this project aims to go deeper. What happens if we compare these AI-generated paintings with paintings from the Renaissance and Baroque based on the same biblical images. That my enlighten both the function of AI-image generation and the workflow of masterpieces of the history of art. Are the skyscrapers in the background at one of the AI-generated images of the Last Supper in essence different from the mediaeval castles we find in the background in Renaissance paintings? Or what does a dog do in some of those Renaissance pictures of the Last Supper, whereas there is no explicit clue in the text to include this animal? 

For the evaluation of the paintings we need your help with this exciting project. To complement or correct the automatic object recognition, we have made a survey of some Renaissance and Baroque paintings with questions to identify the persons, objects and other features in the paintings. We have five topic, each of them with five paintings. Filling in the survey for one topic (five paintings) may take 15 to 30 minutes. If you can only one or two of these surveys it is completely fine. If you wish you can also be kept informed about the outcomes of the project. Here are the links to the surveys:

Expulsion from paradise

https://form.jotform.com/230525328869363 

Tower of Babel

https://form.jotform.com/230873630408354 

Binding of Isaac

https://form.jotform.com/230937362199364 

Moses Found

https://form.jotform.com/230873548200352 

The Last Supper

https://form.jotform.com/230936974087368 

Your help is highly appreciated!

Kind regards,

Willem van Peursen