How to Train your Text-to-Image Model: Evaluating Design Choices for Synthetic Training Captions
This work addresses the need for practical guidance on training data strategies in text-to-image generation, offering incremental insights for researchers and practitioners.
The study tackled the problem of how synthetic caption design choices affect text-to-image model performance, finding that dense captions improve text alignment but reduce aesthetics and diversity, while randomized-length captions balance these aspects without losing diversity.
Training data is at the core of any successful text-to-image models. The quality and descriptiveness of image text are crucial to a model's performance. Given the noisiness and inconsistency in web-scraped datasets, recent works shifted towards synthetic training captions. While this setup is generally believed to produce more capable models, current literature does not provide any insights into its design choices. This study closes this gap by systematically investigating how different synthetic captioning strategies impact the downstream performance of text-to-image models. Our experiments demonstrate that dense, high-quality captions enhance text alignment but may introduce trade-offs in output aesthetics and diversity. Conversely, captions of randomized lengths yield balanced improvements across aesthetics and alignment without compromising sample diversity. We also demonstrate that varying caption distributions introduce significant shifts in the output bias of a trained model. Our findings underscore the importance of caption design in achieving optimal model performance and provide practical insights for more effective training data strategies in text-to-image generation.