DALL-Eval: Probing the Reasoning Skills and Social Biases of Text-to-Image Generation Models

TL;DR

This paper evaluates the reasoning abilities and social biases of state-of-the-art text-to-image models, revealing significant gaps in visual reasoning skills and the presence of biases learned from web data.

Contribution

It introduces PaintSkills, a diagnostic dataset for assessing visual reasoning, and provides a comprehensive analysis of biases in recent models.

Findings

Models perform poorly on object counting and spatial reasoning tasks.

Recent models exhibit gender and skin tone biases learned from web data.

The study highlights the need for improved reasoning and bias mitigation in text-to-image models.

Abstract

Recently, DALL-E, a multimodal transformer language model, and its variants, including diffusion models, have shown high-quality text-to-image generation capabilities. However, despite the realistic image generation results, there has not been a detailed analysis of how to evaluate such models. In this work, we investigate the visual reasoning capabilities and social biases of different text-to-image models, covering both multimodal transformer language models and diffusion models. First, we measure three visual reasoning skills: object recognition, object counting, and spatial relation understanding. For this, we propose PaintSkills, a compositional diagnostic evaluation dataset that measures these skills. Despite the high-fidelity image generation capability, a large gap exists between the performance of recent models and the upper bound accuracy in object counting and spatial relation understanding skills. Second, we assess the gender and skin tone biases by measuring the gender/skin tone distribution of generated images across various professions and attributes. We demonstrate that recent text-to-image generation models learn specific biases about gender and skin tone from web image-text pairs. We hope our work will help guide future progress in improving text-to-image generation models on visual reasoning skills and learning socially unbiased representations. Code and data: https://github.com/j-min/DallEval