Exploiting Relationship for Complex-scene Image Generation

Authors

  • Tianyu Hua JD AI Research
  • Hongdong Zheng JD AI Research
  • Yalong Bai JD AI Research
  • Wei Zhang JD AI Research
  • Xiao-Ping Zhang Ryerson University
  • Tao Mei JD AI Research

Keywords:

Multi-modal Vision, Visual Reasoning & Symbolic Representations, Language and Vision, Multimodal Learning

Abstract

The significant progress on Generative Adversarial Networks (GANs) has facilitated realistic single-object image generation based on language input. However, complex-scene generation (with various interactions among multiple objects) still suffers from messy layouts and object distortions, due to diverse configurations in layouts and appearances. Prior methods are mostly object-driven and ignore their inter-relations that play a significant role in complex-scene images. This work explores relationship-aware complex-scene image generation, where multiple objects are inter-related as a scene graph. With the help of relationships, we propose three major updates in the generation framework. First, reasonable spatial layouts are inferred by jointly considering the semantics and relationships among objects. Compared to standard location regression, we show relative scales and distances serve a more reliable target. Second, since the relations between objects have significantly influenced an object's appearance, we design a relation-guided generator to generate objects reflecting their relationships. Third, a novel scene graph discriminator is proposed to guarantee the consistency between the generated image and the input scene graph. Our method tends to synthesize plausible layouts and objects, respecting the interplay of multiple objects in an image. Experimental results on Visual Genome and HICO-DET datasets show that our proposed method significantly outperforms prior arts in terms of IS and FID metrics. Based on our user study and visual inspection, our method is more effective in generating logical layout and appearance for complex-scenes.

Downloads

Published

2021-05-18

How to Cite

Hua, T., Zheng, H., Bai, Y., Zhang, W., Zhang, X.-P., & Mei, T. (2021). Exploiting Relationship for Complex-scene Image Generation. Proceedings of the AAAI Conference on Artificial Intelligence, 35(2), 1584-1592. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/16250

Issue

Section

AAAI Technical Track on Computer Vision I