GraphDeco

Multi-view Relighting Using a Geometry-Aware Network
Presentation | Team members | Collaborations | Publications | Job offers | Contact

 

Multi-view Relighting Using a Geometry-Aware Network

ACM Transactions on Graphics (SIGGRAPH Conference Proceedings), Volume 38, Number 4 - July 2019
Download the publication : Multi-view-Relighting.pdf [46.7Mo]  
We propose the first learning-based algorithm that can relight images in a plausible and controllable manner given multiple views of an outdoor scene. In particular, we introduce a geometry-aware neural network that utilizes multiple geometry cues (normal maps, specular direction, etc.) and source and target shadow masks computed from a noisy proxy geometry obtained by multi-view stereo. Our model is a three-stage pipeline: two subnetworks refine the source and target shadow masks, and a third performs the final relighting. Furthermore, we introduce a novel representation for the shadow masks, which we call RGB shadow images. They reproject the colors from all views into the shadowed pixels and enable our network to cope with inacuraccies in the proxy and the non-locality of the shadow casting interactions. Acquiring large-scale multi-view relighting datasets for real scenes is challenging, so we train our network on photorealistic synthetic data. At train time, we also compute a noisy stereo-based geometric proxy, this time from the synthetic renderings. This allows us to bridge the gap between the real and synthetic domains. Our model generalizes well to real scenes. It can alter the illumination of drone footage, image-based renderings, textured mesh reconstructions, and even internet photo collections.

Images and movies

 

See also




See also the project webpage.

Acknowledgements and Funding

The authors thank G. Kopanas, L. Boiron and S. Morgenthaler for the development of the 3DSMax to Mitsuba exporter and ground truth rendering system. Thanks to A. Bousseau and K. Sunkavalli for proofreading earlier drafts. Funding was provided by the European Commission grants EMOTIVE H2020 project No. 727188 and ERC Advanced Grant FUNGRAPH (No. 788065, http://fungraph.inria.fr). Drone footage copyright Drones Yucat√°n and Namyeska - info@namyeska.com.

BibTex references

@Article{PGZED19,
  author       = "Philip, Julien and Gharbi, Micha{\"e}l and Zhou, Tinghui and Efros, Alexei and Drettakis, George",
  title        = "Multi-view Relighting Using a Geometry-Aware Network",
  journal      = "ACM Transactions on Graphics (SIGGRAPH Conference Proceedings)",
  number       = "4",
  volume       = "38",
  month        = "July",
  year         = "2019",
  url          = "http://www-sop.inria.fr/reves/Basilic/2019/PGZED19"
}

Other publications in the database

» Julien Philip
» Alexei Efros
» George Drettakis