首页|CRD-CGAN:category-consistent and relativistic constraints for diverse text-to-image generation

CRD-CGAN:category-consistent and relativistic constraints for diverse text-to-image generation

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
Generating photo-realistic images from a text description is a challenging problem in computer vision.Previous works have shown promising performance to generate synthetic images conditional on text by Generative Adversarial Networks(GANs).In this paper,we focus on the category-consistent and relativistic diverse constraints to optimize the diversity of synthetic images.Based on those constraints,a category-consistent and relativistic diverse conditional GAN(CRD-CGAN)is proposed to synthesize K photo-realistic images simultaneously.We use the attention loss and diversity loss to improve the sensitivity of the GAN to word attention and noises.Then,we employ the relativistic conditional loss to estimate the probability of relatively real or fake for synthetic images,which can improve the performance of basic conditional loss.Finally,we introduce a category-consistent loss to alleviate the over-category issues between K synthetic images.We evaluate our approach using the Caltech-UCSD Birds-200-2011,Oxford 102 flower and MS COCO 2014 datasets,and the extensive experiments demonstrate superiority of the proposed method in comparison with state-of-the-art methods in terms of photorealistic and diversity of the generated synthetic images.

text-to-imagediverse conditional GANrelativi-stic category-consistent

Tao HU、Chengjiang LONG、Chunxia XIAO

展开 >

College of Intelligent Systems Science and Engineering,Hubei Minzu University,Enshi 445000,China

School of Computer Science,Wuhan University,Wuhan 430072,China

Key Laboratory of Performing Art Equipment & System Technology,Ministry of Culture and Tourism,Beijing 100007,China

Meta Reality Labs,Burlingame,CA,94010,USA

展开 >

National Natural Science Foundation of ChinaNational Natural Science Foundation of ChinaNational Cultural and Tourism Science and Technology Innovation ProjectTraining Program of High Level Scientific Research Achievements of Hubei Minzu University

61972298619620192021064PY22011

2024

10.1007/s11704-022-2385-x

计算机科学前沿
高等教育出版社

计算机科学前沿

CSTPCDEI
影响因子：0.303
ISSN：2095-2228
年,卷(期)：2024.18(1)
  • 69