HDR Environment Map Estimation for Real-Time Augmented Reality
Abstract
We present a method to estimate an HDR environment map from a narrow field-of-view LDR camera image in real-time. This enables perceptually appealing reflections and shading on virtual objects of any material finish, from mirror to diffuse, rendered into a real physical environment using augmented reality. Our method is based on our efficient convolutional neural network architecture, EnvMapNet, trained end-to-end with two novel losses, ProjectionLoss for the generated image, and ClusterLoss for adversarial training. Through qualitative and quantitative comparison to state-of-the-art methods, we demonstrate that our algorithm reduces the directional error of estimated light sources by more than 50%, and achieves 3.7 times lower Frechet Inception Distance (FID). We further showcase a mobile application that is able to run our neural network model in under 9 ms on an iPhone XS, and render in real-time, visually coherent virtual objects in previously unseen real-world environments.
Cite
@article{arxiv.2011.10687,
title = {HDR Environment Map Estimation for Real-Time Augmented Reality},
author = {Gowri Somanath and Daniel Kurz},
journal= {arXiv preprint arXiv:2011.10687},
year = {2021}
}
Comments
Supplementary video at https://docs-assets.developer.apple.com/ml-research/papers/hdr-environment-map.mp4 Code at https://github.com/apple/ml-envmapnet Accepted to CVPR 2021