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

In this paper, we address the limitations of existing text-to-image diffusionmodels in generating demographically fair results when given human-relateddescriptions. These models often struggle to disentangle the target languagecontext from sociocultural biases, resulting in biased image generation. Toovercome this challenge, we propose Fair Mapping, a general, model-agnostic,and lightweight approach that modifies a pre-trained text-to-image model bycontrolling the prompt to achieve fair image generation. One key advantage ofour approach is its high efficiency. The training process only requiresupdating a small number of parameters in an additional linear mapping network.This not only reduces the computational cost but also accelerates theoptimization process. We first demonstrate the issue of bias in generatedresults caused by language biases in text-guided diffusion models. Bydeveloping a mapping network that projects language embeddings into an unbiasedspace, we enable the generation of relatively balanced demographic resultsbased on a keyword specified in the prompt. With comprehensive experiments onface image generation, we show that our method significantly improves imagegeneration performance when prompted with descriptions related to human faces.By effectively addressing the issue of bias, we produce more fair and diverseimage outputs. This work contributes to the field of text-to-image generationby enhancing the ability to generate images that accurately reflect theintended demographic characteristics specified in the text.

Fair Text-to-Image Diffusion via Fair Mapping