On the Fairness, Diversity and Reliability of Text-to-Image Generative Models
This work addresses critical issues of bias and reliability in text-to-image models, which is important for developers and users in AI ethics, but it is incremental as it builds on existing evaluation methods without introducing a new paradigm.
The paper tackles the problem of assessing fairness, diversity, and reliability in text-to-image generative models by proposing an evaluation framework that analyzes responses to perturbations in embedding space, identifying inputs that trigger unreliable or biased behavior, with results including metrics for generative diversity and fairness under low guidance setups.
The rapid proliferation of multimodal generative models has sparked critical discussions on their reliability, fairness and potential for misuse. While text-to-image models excel at producing high-fidelity, user-guided content, they often exhibit unpredictable behaviors and vulnerabilities that can be exploited to manipulate class or concept representations. To address this, we propose an evaluation framework to assess model reliability by analyzing responses to global and local perturbations in the embedding space, enabling the identification of inputs that trigger unreliable or biased behavior. Beyond social implications, fairness and diversity are fundamental to defining robust and trustworthy model behavior. Our approach offers deeper insights into these essential aspects by evaluating: (i) generative diversity, measuring the breadth of visual representations for learned concepts, and (ii) generative fairness, which examines the impact that removing concepts from input prompts has on control, under a low guidance setup. Beyond these evaluations, our method lays the groundwork for detecting unreliable, bias-injected models and tracing the provenance of embedded biases. Our code is publicly available at https://github.com/JJ-Vice/T2I_Fairness_Diversity_Reliability. Keywords: Fairness, Reliability, AI Ethics, Bias, Text-to-Image Models