Skip to main content
Top

Generative Multiplane Images: Making a 2D GAN 3D-Aware

  • 2022
  • OriginalPaper
  • Chapter
Published in:

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

The chapter delves into the development of a novel method to make a 2D GAN 3D-aware by introducing an additional branch for generating alpha maps conditioned on depths and conditioning the discriminator on pose. This approach, termed 'Generative Multiplane Images' (GMPI), ensures view-consistency and efficient rendering, demonstrating significant improvements in 3D-awareness with minimal adjustments to the existing GAN architecture. The method is validated on three high-resolution datasets, showcasing its effectiveness and efficiency compared to state-of-the-art methods. The study also highlights the limitations of traditional 2D GAN evaluation metrics and proposes more suitable metrics for evaluating 3D-aware generative models.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/978-3-031-20065-6_2.
X. Zhao—Work done as part of an internship at Apple.

Not a customer yet? Then find out more about our access models now:

Individual Access

Start your personal individual access now. Get instant access to more than 164,000 books and 540 journals – including PDF downloads and new releases.

Starting from 54,00 € per month!    

Get access

Access for Businesses

Utilise Springer Professional in your company and provide your employees with sound specialist knowledge. Request information about corporate access now.

Find out how Springer Professional can uplift your work!

Contact us now
Title
Generative Multiplane Images: Making a 2D GAN 3D-Aware
Authors
Xiaoming Zhao
Fangchang Ma
David Güera
Zhile Ren
Alexander G. Schwing
Alex Colburn
Copyright Year
2022
DOI
https://doi.org/10.1007/978-3-031-20065-6_2
This content is only visible if you are logged in and have the appropriate permissions.
This content is only visible if you are logged in and have the appropriate permissions.
This content is only visible if you are logged in and have the appropriate permissions.

Premium Partner

    Image Credits
    Neuer Inhalt/© ITandMEDIA, Nagarro GmbH/© Nagarro GmbH, AvePoint Deutschland GmbH/© AvePoint Deutschland GmbH, AFB Gemeinnützige GmbH/© AFB Gemeinnützige GmbH, USU GmbH/© USU GmbH, Ferrari electronic AG/© Ferrari electronic AG