DF-GAN: A Simple and Effective Baseline for Text-to-Image Synthesis

TLDR

Text‑to‑image synthesis is challenging, and current GANs suffer from generator entanglement, limited supervision due to extra networks, and costly cross‑modal attention fusion. The authors propose DF‑GAN, a simpler yet more effective text‑to‑image GAN. DF‑GAN uses a one‑stage backbone that directly generates high‑resolution images, a target‑aware discriminator with matching‑aware gradient penalty and one‑way output to improve semantic consistency, and a deep fusion block that fully integrates text and visual features; the code is available online. Compared with state‑of‑the‑art methods, DF‑GAN is simpler, more efficient, and achieves better performance on widely used datasets.

Abstract

Synthesizing high-quality realistic images from text descriptions is a challenging task. Existing text-to-image Generative Adversarial Networks generally employ a stacked architecture as the backbone yet still remain three flaws. First, the stacked architecture introduces the entanglements between generators of different image scales. Second, existing studies prefer to apply and fix extra networks in adversarial learning for text-image semantic consistency, which limits the supervision capability of these networks. Third, the cross-modal attention-based text-image fusion that widely adopted by previous works is limited on several special image scales because of the computational cost. To these ends, we propose a simpler but more effective Deep Fusion Generative Adversarial Networks (DF-GAN). To be specific, we propose: (i) a novel one-stage text-to-image backbone that directly synthesizes high-resolution images without entanglements between different generators, (ii) a novel Target-Aware Discriminator composed of Matching-Aware Gradient Penalty and One-Way Output, which enhances the text-image semantic consistency without introducing extra networks, (iii) a novel deep text-image fusion block, which deepens the fusion process to make a full fusion between text and visual features. Compared with current state-of-the-art methods, our proposed DF-GAN is simpler but more efficient to synthesize realistic and text-matching images and achieves better performance on widely used datasets. Code is available at https://github.com/tobran/DF-GAN.