New journal paper at SPRINGER
On March 11th, 2020 a new paper was published which uses neural networks to reconstruct the lighting of the environment.
A journal paper with the title "Environment Estimation for Glossy Reflections in Mixed Reality Applications Using a Neural Network" was successfully placed in the journal Transactions on Computational Science XXXVI at Springer Verlag. This is an extension of the paper at Cyberworlds 2018. The authors of the paper are Tobias Schwandt, Christian Kunert and Wolfgang Broll.
The content of the paper is the estimation of the ambient lighting using previously seen image information in combination with a neural network. A conventional smartphone is used, which creates 360 ° lighting via the camera and the movement data. Missing parts that have not yet been seen by the camera are estimated using the neural network. Further information and the paper can be found on the following links: