Towards Understanding and Quantifying Uncertainty for Text-to-Image Generation

Test Time Uncertainty Quantification Generated Images <prompts> Training Dataset concept 1 concept 2 concept 3 Bias Extraction Prompt-based UNCertainty Estimation for Text-to-Image (T2I) Generation Deep Fake Prevention Copyright Prevention Figure 1. Examples of Applications for Uncertainty Quantification in Text-to-Image Generation. Text-to-image generation models may exhibit uncertainty, and that need to be quantified since it can provide insights into the model's training dataset, aiding in deepfake prevention, detecting model biases, and protecting copyrighted content from unauthorized generation.