# Neural Global Illumination: Interactive Indirect Illumination Prediction Under Dynamic Area Lights Duan Gao, Haoyuan Mu, Kun Xu · 2023 · *IEEE Transactions on Visualization and Computer Graphics* [file](https://files.dgkn.uk/papers/2023-gao-neural-global-illumination-interactive-indirect-illumination.pdf) · [DOI](https://doi.org/10.1109/TVCG.2022.3209963) ## Abstract We propose neural global illumination, a novel method for fast rendering full global illumination in static scenes with dynamic viewpoint and area lighting. The key idea of our method is to utilize a deep rendering network to model the complex mapping from each shading point to global illumination. To efficiently learn the mapping, we propose a neural-network-friendly input representation including attributes of each shading point, viewpoint information, and a combinational lighting representation that enables high-quality fitting with a compact neural network. To synthesize high-frequency global illumination effects, we transform the low-dimension input to higher-dimension space by positional encoding and model the rendering network as a deep fully-connected network. Besides, we feed a screen-space neural buffer to our rendering network to share global information between objects in the screen-space to each shading point. We have demonstrated our neural global illumination method in rendering a wide variety of scenes exhibiting complex and all-frequency global illumination effects such as multiple-bounce glossy interreflection, color bleeding, and caustics. ## My notes - Trained on a scene with random placed area lights, used encoded light, etc information and some radiance cues (rendered the view with 4 different materials). Used CNN to encode, and MLP to return a rendered image. Explains each parameter or architecture selection during training, it may become useful for that. Paper states that the results are not as good in the views where no training data is available, so probably the model memorizes (speculation by me). For training they used 5000 images with random camera and area light locations. They stated the method is not good for transparent materials. But the biggest drawback is it is trained on the scene they compared the results in real time.