paper-with-me

홈 › Papers

Can we Debias Social Stereotypes in AI-Generated Images? Examining Text-to-Image Outputs and User Perceptions

2025-05-27 · Saharsh Barve, Andy Mao, Jiayue Melissa Shi, Prerna Juneja, Koustuv Saha

Recent advances in generative AI have enabled visual content creation through text-to-image (T2I) generation. However, despite their creative potential, T2I models often replicate and amplify societal stereotypes -- particularly those related to gender, race, and culture -- raising important ethical concerns. This paper proposes a theory-driven bias detection rubric and a Social Stereotype Index (SSI) to systematically evaluate social biases in T2I outputs. We audited three major T2I model outputs -- DALL-E-3, Midjourney-6.1, and Stability AI Core -- using 100 queries across three categories -- geocultural, occupational, and adjectival. Our analysis reveals that initial outputs are prone to include stereotypical visual cues, including gendered professions, cultural markers, and western beauty norms. To address this, we adopted our rubric to conduct targeted prompt refinement using LLMs, which significantly reduced bias -- SSI dropped by 61% for geocultural, 69% for occupational, and 51% for adjectival queries. We complemented our quantitative analysis through a user study examining perceptions, awareness, and preferences around AI-generated biased imagery. Our findings reveal a key tension -- although prompt refinement can mitigate stereotypes, it can limit contextual alignment. Interestingly, users often perceived stereotypical images to be more aligned with their expectations. We discuss the need to balance ethical debiasing with contextual relevance and call for T2I systems that support global diversity and inclusivity while not compromising the reflection of real-world social complexity.

📄 PDF Abstract BibTeX arXiv:2505.20692

Code (0)

등록된 구현이 없습니다.

Tasks

Bias Detection

Similar Papers 제목 키워드 기반

Social-Group-Agnostic Word Embedding Debiasing via the Stereotype Content Model

2022-10-11 · Ali Omrani, Brendan Kennedy, Mohammad Atari, Morteza Dehghani

Existing word embedding debiasing methods require social-group-specific word pairs (e.g., "man"-"woman") for each social attribute (e.g., gender), which cannot be used to mitigate bias for other social groups, making the…

Attribute

Debiasing Pretrained Text Encoders by Paying Attention to Paying Attention

2021-09-29 · Yacine GACI, Boualem Benatallah, Fabio Casati, Khalid Benabdeslem

Recent studies in fair Representation Learning have observed a strong inclination for natural language processing (NLP) models to exhibit discriminatory stereotypes across gender, religion, race and many such social cons…

FairnessRepresentation LearningSentenceWord Embeddings

Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypes

2024-02-03 · Isabel O. Gallegos, Ryan A. Rossi, Joe Barrow, Md Mehrab Tanjim 외

Large language models (LLMs) have shown remarkable advances in language generation and understanding but are also prone to exhibiting harmful social biases. While recognition of these behaviors has generated an abundance…

Text GenerationZero-Shot Learning

Queer People are People First: Deconstructing Sexual Identity Stereotypes in Large Language Models

2023-06-30 · Harnoor Dhingra, Preetiha Jayashanker, Sayali Moghe, Emma Strubell

Large Language Models (LLMs) are trained primarily on minimally processed web text, which exhibits the same wide range of social biases held by the humans who created that content. Consequently, text generated by LLMs ca…

Sentence

Anti-stereotypical Predictive Text Suggestions Do Not Reliably Yield Anti-stereotypical Writing

2024-09-30 · Connor Baumler, Hal Daumé III

AI-based systems such as language models can replicate and amplify social biases reflected in their training data. Among other questionable behavior, this can lead to LM-generated text--and text suggestions--that contain…

Language ModelingLanguage Modelling