Where Do Images Come From? Analyzing Captions to Geographically Profile Datasets
This work addresses the problem of geographic bias in AI training data for researchers and practitioners, highlighting significant under-representation of certain regions, which is incremental as it builds on prior concerns about dataset representativeness.
The study analyzed the geographic origins of images in large-scale multimodal datasets by extracting location information from captions, finding that 48.0% of samples come from the US, UK, and Canada, while South America and Africa are severely under-represented at 1.8% and 3.8%, respectively, and showing a strong correlation (ρ=0.82) between a country's GDP and its representation.
Recent studies show that text-to-image models often fail to generate geographically representative images, raising concerns about the representativeness of their training data and motivating the question: which parts of the world do these training examples come from? We geographically profile large-scale multimodal datasets by mapping image-caption pairs to countries based on location information extracted from captions using LLMs. Studying English captions from three widely used datasets (Re-LAION, DataComp1B, and Conceptual Captions) across $20$ common entities (e.g., house, flag), we find that the United States, the United Kingdom, and Canada account for $48.0\%$ of samples, while South American and African countries are severely under-represented with only $1.8\%$ and $3.8\%$ of images, respectively. We observe a strong correlation between a country's GDP and its representation in the data ($ρ= 0.82$). Examining non-English subsets for $4$ languages from the Re-LAION dataset, we find that representation skews heavily toward countries where these languages are predominantly spoken. Additionally, we find that higher representation does not necessarily translate to greater visual or semantic diversity. Finally, analyzing country-specific images generated by Stable Diffusion v1.3 trained on Re-LAION, we show that while generations appear realistic, they are severely limited in their coverage compared to real-world images.