2 datasets found

The dataset used in the paper is a binary preference dataset from heterogeneous annotators.

Text-to-image generative models have recently attracted considerable interest, enabling the synthesis of high-quality images from textual prompts. However, these models often...