Education Research Current About VU Amsterdam NL
Login as
Prospective student Student Employee
Bachelor Master VU for Professionals
Exchange programme VU Amsterdam Summer School Honours programme VU-NT2 Semester in Amsterdam
PhD at VU Amsterdam Research highlights Prizes and distinctions
Research institutes Our scientists Research Impact Support Portal Creating impact
News Events calendar Biodiversity at VU Amsterdam
Israël and Palestinian regions Culture on campus
Practical matters Mission and core values Entrepreneurship on VU Campus
Governance Partnerships Alumni University Library Working at VU Amsterdam
Sorry! De informatie die je zoekt, is enkel beschikbaar in het Engels.
This programme is saved in My Study Choice.
Something went wrong with processing the request.
Something went wrong with processing the request.

Seeing The Words: Evaluating AI-generated Biblical Art

This research project started with a project by Hidde Makimei, co-supervised by Shuai Wang and Willem van Peursen. This pilot project involved (a) the generation of a large dataset of over 7K images with AI tools for T2I (text-to-image) generation, using biblical text as prompts and (b) the evaluation of these images with multiple neural network-based tools.

The T2I tools used were DALL-E, MidJourney, and various versions of Stable Diffusion. The prompts used in this pilot were the biblical passages on the expulsion from Paradise (Genesis 4:23–24), the Tower of Babel (Genesis 11:1–9), the binding of Isaac (Genesis 22:9–14), Moses Found (Exodus 2:5–9) and the Last Supper 14:12–25). The generated images were further compared (both manually and automatically) with Renaissance and Baroque paintings based on the same episodes. In this way we explored the complex interaction between text, imagination and conventional representations and AI and evaluated the various aspects of AI tools for T2I generation.

A challenge that emerged from the first project were unexpected elements in the images that seemed to be alien to the biblical passages. Should we evaluate them in terms of inaccuracy, or rather in terms of creativity? This motivated a follow-up in Samuel Ebenezer Entsua-Mensah’s project (supervised by Van Peursen), In this second project we select only one of the T2I tools of the first project, namely Midjourney (which performed best in the first project), and one of the biblical episodes, namely Moses Found. An additional element in this project is the comparison with images that emerge from a Google search for images based on the same passage. The leading question of this project is: Now that the performance of models such as MidJourney has increased considerably and the results resemble so much what can be found by available images on the internet, is there any room left for creativity or “augmented imagination” (a term used by David Holz the founder of MidJourney) given that the models perform so well in mirroring what humans can do.

A PhD project by Anny Valentia (supervised by Marius Dorobantu and Willem van Peursen) focuses on two biblical passages from the Bible: the Tower of Babel and Ezekiel’s opening visions. The first are typical of the sober style of biblical narrative as described by Eric Auerbach, leaving all visual details to the reader’s imagination, the second has a rich descriptive style, detailing all kinds of visual details, which is further developed in apocalyptic literature.

More about this Research Project

Start/end Date

Ongoing since 1 January 2023

Team

Leader: Willem Th. van Peursen
Team: Shuai Wang, Hidde Makimei (2023–2024), Marius Dorobantu (from 2024), Anny Valentina (from 2024), Samuel Ebenezer Entsua-Mensah (from 2025).

Websites

The Dataset can be found on the DANS SSH Data Station.

Source code

Published version of the source code

A draft version of the corresponding paper is available at Hidde Makimei, Shuai Wang, Shuai, and Willem van Peursen, “Seeing the Words: Evaluating AI-generated Biblical Art” 

Quick links

Homepage Culture on campus VU Sports Centre Dashboard

Study

Academic calendar Study guide Timetable Canvas

Featured

VUfonds VU Magazine Ad Valvas Digital accessibility

About VU

Contact us Working at VU Amsterdam Faculties Divisions
Privacy Disclaimer Safety Web Colophon Cookie Settings Web Archive

Copyright © 2025 - Vrije Universiteit Amsterdam