Capsule based image synthesis for interior design effect rendering

Fei Yang, Zheng Lu, Guoping Qiu, Jing Lin, Qian Zhang

Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

3 Citations (Scopus)


Effect rendering that renders 3D model to 2D images with various coloring and lighting effects, is an important step in home interior design. Traditional way of manual rendering using professional software is very labor intensive and time consuming. In this paper, we present a novel capsule based conditional generative adversarial network that can automatically synthesize an indoor image with realistic and aesthetically pleasing rendering effect from a given plain image rendered without any effects from a interior designed 3D model. By adapting capsule blocks in both generator and discriminator and a novel multi-way loss function inside discriminator, our framework is able to generate more realistic rendering effect at both detail and global levels. In addition, a novel line preservation loss is introduced not only to help preserve the properties that are independent of lighting effect, but also improves the lighting effect along those lines. We apply our technique on a dataset specially prepared for interior design effect rendering and systematically compare our approach with multiple state-of-the-art methods.

Original languageEnglish
Title of host publicationComputer Vision – ACCV 2018 - 14th Asian Conference on Computer Vision, Revised Selected Papers
EditorsC.V. Jawahar, Greg Mori, Hongdong Li, Konrad Schindler
PublisherSpringer Verlag
Number of pages16
ISBN (Print)9783030208721
Publication statusPublished - 2019
Event14th Asian Conference on Computer Vision, ACCV 2018 - Perth, Australia
Duration: 2 Dec 20186 Dec 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11365 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference14th Asian Conference on Computer Vision, ACCV 2018


  • Capsule
  • GAN
  • Image synthesize
  • Rendering

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Capsule based image synthesis for interior design effect rendering'. Together they form a unique fingerprint.

Cite this