AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks
It addresses the challenge of generating high-quality, detailed images from text for applications in AI and computer vision, representing a significant advance over prior methods.
The paper tackles the problem of fine-grained text-to-image generation by proposing AttnGAN, which uses attention mechanisms to synthesize detailed images from text descriptions, resulting in a 14.14% improvement in inception score on the CUB dataset and 170.25% on COCO.
In this paper, we propose an Attentional Generative Adversarial Network (AttnGAN) that allows attention-driven, multi-stage refinement for fine-grained text-to-image generation. With a novel attentional generative network, the AttnGAN can synthesize fine-grained details at different subregions of the image by paying attentions to the relevant words in the natural language description. In addition, a deep attentional multimodal similarity model is proposed to compute a fine-grained image-text matching loss for training the generator. The proposed AttnGAN significantly outperforms the previous state of the art, boosting the best reported inception score by 14.14% on the CUB dataset and 170.25% on the more challenging COCO dataset. A detailed analysis is also performed by visualizing the attention layers of the AttnGAN. It for the first time shows that the layered attentional GAN is able to automatically select the condition at the word level for generating different parts of the image.