Zero-Shot Text-to-Image Generation
This addresses the problem of generating images from text without domain-specific training, offering a more generalizable solution for AI applications.
The paper tackles text-to-image generation by proposing a simple transformer-based approach that models text and image tokens autoregressively as a single data stream, achieving competitive performance with previous domain-specific models in a zero-shot setting.
Text-to-image generation has traditionally focused on finding better modeling assumptions for training on a fixed dataset. These assumptions might involve complex architectures, auxiliary losses, or side information such as object part labels or segmentation masks supplied during training. We describe a simple approach for this task based on a transformer that autoregressively models the text and image tokens as a single stream of data. With sufficient data and scale, our approach is competitive with previous domain-specific models when evaluated in a zero-shot fashion.