Efficient Neural Architecture for Text-to-Image Synthesis

By Douglas M. Souza, Jonatas Wehrmann, Duncan Dubugras Alcoba Ruiz. 2020

25 citations recorded by the source corpus.

Text-to-image synthesis is the task of generating images from text descriptions. Image generation, by itself, is a challenging task. When we combine image generation and text, we bring complexity to a new level: we need to combine data from two different modalities. Most of recent works in text-to-image synthesis follow a similar approach when it comes to neural architectures. Due to aforementioned difficulties, plus the inherent difficulty of training GANs at high resolutions, most methods have adopted a multi-stage training strategy. In this paper we shift the architectural paradigm currently used in text-to-image methods and show that an effective neural architecture can achieve state-of-the-art performance using a single stage training with a single generator and a single discriminator. We do so by applying deep residual networks along with a novel sentence interpolation strategy that enables learning a smooth conditional space. Finally, our work points a new direction for text-to-image research, which has not experimented with novel neural architectures recently.

Subfields: Computer Vision and Pattern Recognition, Generative Models, Multimodal Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Generative Models

In Multimodal Learning

Citations in the corpus

Adversarial Text-to-Image Synthesis: A Review (2021, methodology): “In [34], the authors adapted BigGAN [84], an architecture that has previously presented a new state-of-the-art on ImageNet conditioned on class labels, for T2I synthesis.”