CVLGDec 13, 2023

Stable Rivers: A Case Study in the Application of Text-to-Image Generative Models for Earth Sciences

arXiv:2312.07833v14 citationsh-index: 3Earth Surface Processes and Landforms
Originality Synthesis-oriented
AI Analysis

It addresses domain-specific biases in text-to-image models for earth sciences, which is an incremental step beyond existing social and cultural bias research.

The study evaluated biases in Stable Diffusion's training data for fluvial geomorphology, finding over-representation of scenic locations and under-representation of morphological terms, but showed that careful prompting could generate photorealistic synthetic river images.

Text-to-image (TTI) generative models can be used to generate photorealistic images from a given text-string input. These models offer great potential to mitigate challenges to the uptake of machine learning in the earth sciences. However, the rapid increase in their use has raised questions about fairness and biases, with most research to-date focusing on social and cultural areas rather than domain-specific considerations. We conducted a case study for the earth sciences, focusing on the field of fluvial geomorphology, where we evaluated subject-area specific biases in the training data and downstream model performance of Stable Diffusion (v1.5). In addition to perpetuating Western biases, we found that the training data over-represented scenic locations, such as famous rivers and waterfalls, and showed serious under- and over-representation of many morphological and environmental terms. Despite biased training data, we found that with careful prompting, the Stable Diffusion model was able to generate photorealistic synthetic river images reproducing many important environmental and morphological characteristics. Furthermore, conditional control techniques, such as the use of condition maps with ControlNet were effective for providing additional constraints on output images. Despite great potential for the use of TTI models in the earth sciences field, we advocate for caution in sensitive applications, and advocate for domain-specific reviews of training data and image generation biases to mitigate perpetuation of existing biases.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes