Editable Indoor Lighting Estimation
Henrique Weber     Mathieu Garon     Jean-François Lalonde    

Accepted at European Conference on Computer Vision (ECCV), 2022!


We present a method for estimating lighting from a single perspective image of an indoor scene. Previous methods for predicting indoor illumination usually focus on either simple, parametric lighting that lack realism, or on richer representations that are difficult or even impossible to understand or modify after prediction. We propose a pipeline that estimates a parametric light that is easy to edit and allows renderings with strong shadows, alongside with a non-parametric texture with high-frequency information necessary for realistic rendering of specular objects. Once estimated, the predictions obtained with our model are interpretable and can easily be modified by an artist/user with a few mouse clicks. Quantitative and qualitative results show that our approach makes indoor lighting estimation easier to handle by a casual user, while still producing competitive results.

Paper (Arxiv)

Henrique Weber, Mathieu Garon, Jean-François Lalonde
Editable Indoor Lighting Estimation. European Conference on Computer Vision (ECCV), 2022


Test Set Data and Comparison Code

Test Set Images
Renders with GT envmaps
Scripts to render and generate Table 1 of the paper


poster pdf




Some results with our approach. The first few frames show the initial prediction, and subsequent frames how we can modify the lighting (for example, by changing its azimuth, elevation, or size).


This research was supported by MITACS and the NSERC grant RGPIN-2020-04799. The authors thank Pascal Audet for his help.