English

Attribute-Centric Compositional Text-to-Image Generation

Computer Vision and Pattern Recognition 2023-01-05 v1

Abstract

Despite the recent impressive breakthroughs in text-to-image generation, generative models have difficulty in capturing the data distribution of underrepresented attribute compositions while over-memorizing overrepresented attribute compositions, which raises public concerns about their robustness and fairness. To tackle this challenge, we propose ACTIG, an attribute-centric compositional text-to-image generation framework. We present an attribute-centric feature augmentation and a novel image-free training scheme, which greatly improves model's ability to generate images with underrepresented attributes. We further propose an attribute-centric contrastive loss to avoid overfitting to overrepresented attribute compositions. We validate our framework on the CelebA-HQ and CUB datasets. Extensive experiments show that the compositional generalization of ACTIG is outstanding, and our framework outperforms previous works in terms of image quality and text-image consistency.

Keywords

Cite

@article{arxiv.2301.01413,
  title  = {Attribute-Centric Compositional Text-to-Image Generation},
  author = {Yuren Cong and Martin Renqiang Min and Li Erran Li and Bodo Rosenhahn and Michael Ying Yang},
  journal= {arXiv preprint arXiv:2301.01413},
  year   = {2023}
}
R2 v1 2026-06-28T08:01:53.813Z