Inference Time Debiasing Concepts in Diffusion Models
This addresses bias in AI-generated images for users of diffusion models, offering an accessible, low-overhead solution, though it is incremental as it builds on existing diffusion frameworks.
The authors tackled bias in text-to-image diffusion models by proposing DeCoDi, an inference-time debiasing method that modifies the diffusion process to avoid biased concepts, and demonstrated its effectiveness in reducing gender, ethnicity, and age biases for concepts like nurse and CEO through human evaluation of 1,200 images.
We propose DeCoDi, a debiasing procedure for text-to-image diffusion-based models that changes the inference procedure, does not significantly change image quality, has negligible compute overhead, and can be applied in any diffusion-based image generation model. DeCoDi changes the diffusion process to avoid latent dimension regions of biased concepts. While most deep learning debiasing methods require complex or compute-intensive interventions, our method is designed to change only the inference procedure. Therefore, it is more accessible to a wide range of practitioners. We show the effectiveness of the method by debiasing for gender, ethnicity, and age for the concepts of nurse, firefighter, and CEO. Two distinct human evaluators manually inspect 1,200 generated images. Their evaluation results provide evidence that our method is effective in mitigating biases based on gender, ethnicity, and age. We also show that an automatic bias evaluation performed by the GPT4o is not significantly statistically distinct from a human evaluation. Our evaluation shows promising results, with reliable levels of agreement between evaluators and more coverage of protected attributes. Our method has the potential to significantly improve the diversity of images it generates by diffusion-based text-to-image generative models.