Research map / Image generation / Text-to-image diffusion
Navigating Text-to-Image Generative Bias across Indic Languages
Surbhi Mittal, Arnav Sudan, Mayank Vatsa, Richa Singh, Tamar Glaser, Tal Hassner
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 |
|---|---|---|---|
| prompt in Indic languages instead of English | Stable Diffusion · IndicTTI Assamese | CLGC (%): 64.53 -> 7.29 | Hurt |
| prompt in Indic languages instead of English | Alt Diffusion · IndicTTI Assamese | CLGC (%): 58.16 -> 14.65 | Hurt |
| use Dalle3 instead of open-source models for Indic prompts | text-to-image generation · IndicTTI Hindi | CLGC (%): 7.35 -> 64.16 | Improved |
| use multilingual-trained AltDiffusion instead of Stable Diffusion for Indic prompts | Stable Diffusion · IndicTTI Hindi | CLGC (%): 7.35 -> 16.79 | Improved |
| use multilingual-trained AltDiffusion instead of Stable Diffusion for English prompts | Stable Diffusion · IndicTTI English | CLGC (%): 64.53 -> 58.16 | Hurt |
| Dalle3 on low-resource scripts (Meitei, Ol Chiki, Bodo) | Dalle3 · IndicTTI LGC | LGC (%): 33.06 (English) -> 1.52 Manipuri Meitei, 2.09 Santali | Hurt |
Closest papers by meaning
- IndicVisionBench: Benchmarking Cultural and Multilingual Understanding in VLMs (ICLR 2026 · similarity 0.89)
- MuLan: Adapting Multilingual Diffusion Models for Hundreds of Languages with Negligible Cost (ICML 2025 · similarity 0.87)
- On Fairness of Unified Multimodal Large Language Model for Image Generation (NeurIPS 2025 · similarity 0.86)
- AltDiffusion: A Multilingual Text-to-Image Diffusion Model (AAAI 2024 · similarity 0.86)
- OpenBias: Open-set Bias Detection in Text-to-Image Generative Models (CVPR 2024 · similarity 0.86)
- See It from My Perspective: How Language Affects Cultural Bias in Image Understanding (ICLR 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.
