English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Text-to-image diffusion models have been adopted into key commercialworkflows, such as art generation and image editing. Characterising theimplicit social biases they exhibit, such as gender and racial stereotypes, isa necessary first step in avoiding discriminatory outcomes. While existingstudies on social bias focus on image generation, the biases exhibited inalternate applications of diffusion-based foundation models remainunder-explored. We propose methods that use synthetic images to probe twoapplications of diffusion models, image editing and classification, for socialbias. Using our methodology, we uncover meaningful and significantinter-sectional social biases in \textit{Stable Diffusion}, a state-of-the-artopen-source text-to-image model. Our findings caution against the uninformedadoption of text-to-image foundation models for downstream tasks and services.

Exploring Social Bias in Downstream Applications of Text-to-Image Foundation Models