Deep Shading

Convolutional Neural Networks for Screen-Space Shading

Our project is based on Nalbach et al. 2017 paper. In this project, a set of buffers are provided to a Deep Convolutional Network in order to synthetize differend shading effects (such as Ambient Occlusion, Depth of Field, Global Illumination and Sub-surface Scattering). The set of buffers depends os the shading effect we want to synthetize.

Input Bufers:

Result (Ambient Occlusion):

Requirements for the project

0. Python 3.x
1. Tensorflow 1.10
2. Keras
3. OpenCV 3.4(for loading,resizing images)
4. h5py(for saving trained model)
5. pyexr

Steps to run the repo

1. Clone the repo
   
2. Download the dataset (http://deep-shading-datasets.mpi-inf.mpg.de/)
   
3. Install the requirements
   
4. Generate the .tfrecord for trainning and validation (Use the DataReader.py)
   
5. Run "Shading.py"

References

[1] Oliver Nalbach, Elena Arabadzhiyska, Dushyant Mehta, Hans-Peter Seidel, Tobias Ritschel Deep Shading: Convolutional Neural Networks for Screen-Space Shading to appear in Proc. EGSR 2017