Research map / Image generation / Text-to-image diffusion
When Cars Have Stereotypes: Auditing Demographic Bias in Objects from Text-to-Image Models
Dasol Choi, Jihwan Lee, Minjae Lee, Minsuk Kahng
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add 'avoiding stereotypical styles' suffix (ablation) | GPT, Flux, Qwen · cars and sofas | CDS: -30.9% | Improved |
| add 'avoiding stereotypical styles' suffix (ablation) | GPT, Flux, Qwen · cars and sofas | VAC: +8.4% | Hurt |
| add 'with a wide range of styles' suffix (ablation) | GPT, Flux, Qwen · cars and sofas | CDS: -11.6% | Improved |
| add 'with a wide range of styles' suffix (ablation) | GPT, Flux, Qwen · cars and sofas | VAC: +6.8% | Hurt |
| replace 'for' with 'designed for' (ablation) | original template · cars and sofas | CDS: -10.6% | Improved |
| replace 'for' with 'targeted at' (ablation) | original template · cars and sofas | VAC: +6.6% | Hurt |
| replace 'for' with 'targeted at' (ablation) | original template · cars and sofas | CDS: -0.4% | No clear effect |
| add 'avoiding stereotypical styles' suffix (ablation) | Flux.2 Pro · cars and sofas | CDS and VAC: CDS -34.7%, VAC -5.5% | Improved |
| add 'avoiding stereotypical styles' suffix (ablation) | GPT Image-1 · cars and sofas | CDS and VAC: CDS -32.8%, VAC +9.9% | Mixed |
| add 'avoiding stereotypical styles' suffix (ablation) | Qwen Image · cars and sofas | CDS and VAC: CDS -24.6%, VAC +19.3% | Mixed |
Closest papers by meaning
- OASIS Uncovers: High-Quality T2I Models, Same Old Stereotypes (ICLR 2025 · similarity 0.90)
- Association of Objects May Engender Stereotypes: Mitigating Association-Engendered Stereotypes in Text-to-Image Generation (NeurIPS 2024 · similarity 0.88)
- SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Examples (CVPR 2024 · similarity 0.88)
- Implicit Bias Injection Attacks against Text-to-Image Diffusion Models (CVPR 2025 · similarity 0.86)
- FairImagen: Post-Processing for Bias Mitigation in Text-to-Image Models (NeurIPS 2025 · similarity 0.86)
- Multi-Group Proportional Representations for Text-to-Image Models (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
