AI images pre-test

Autor: Roura, Mireia Triguero
Rok vydání: 2022
Předmět:
DOI: 10.17605/osf.io/r4jeh
Popis: The goal of this online task is to understand how people in different countries perceive AI-generated images of users. 66 images were created using Generated Images AI. The images correspond to 11 "men" and 11 "women" that are the same on everything except for skin tone. skin tones take three possible values according to the AI's parameters (fair, olive and black). These images represent the main instrument in a different project. The goal of this pre-test is to see whether people perceive the skin tone and gender in a way consistent with the AI's parameters on gender and skin tone. Finally, it also intends to understand if people make consistent racial and regional classifications based exclusively on the image.
Databáze: OpenAIRE