Abstract
Over recent years, the use of artificial intelligence (AI) in the field of Art History has garnered growing interest. Many academic publications on this relatively recent topic explore the role of AI in the analysis of huge datasets and
digitised art collections, according to specific research or curatorial questions, while others address AI as a theme or a tool for contemporary artistic practices. This paper presents an alternative approach, considering generative AI as part of an interpretative methodology based on derivative images created with text prompts that specifically request a reinterpretation of a particular artwork, without adding any stylistic or contextual modifiers. Focusing on the iconic Self-Portrait (in a redcoat) by the Portuguese painter Aurélia de Souza, the aim of this study is to discuss how images produced with different text-to-image AI generators may not only illustrate some of the features highlighted in Art History studies, but also foster new questions and readings of the same artwork.
digitised art collections, according to specific research or curatorial questions, while others address AI as a theme or a tool for contemporary artistic practices. This paper presents an alternative approach, considering generative AI as part of an interpretative methodology based on derivative images created with text prompts that specifically request a reinterpretation of a particular artwork, without adding any stylistic or contextual modifiers. Focusing on the iconic Self-Portrait (in a redcoat) by the Portuguese painter Aurélia de Souza, the aim of this study is to discuss how images produced with different text-to-image AI generators may not only illustrate some of the features highlighted in Art History studies, but also foster new questions and readings of the same artwork.
Original language | English |
---|---|
Title of host publication | EVA Berlin 2023 |
Subtitle of host publication | Elektronische Medien & Kunst, Kultur und Historie |
Editors | Dominik Lengyel |
Place of Publication | Berlin |
Publisher | BTU Brandenburgische Technische Universität Cottbus-Senftenberg |
Pages | 286-294 |
Number of pages | 8 |
ISBN (Electronic) | 978-3-88609-891-0 |
Publication status | Published - 2023 |
Event | EVA Berlin 2023 - Berlin, Germany Duration: 29 Nov 2023 → 1 Dec 2023 |
Conference
Conference | EVA Berlin 2023 |
---|---|
Country/Territory | Germany |
City | Berlin |
Period | 29/11/23 → 1/12/23 |