University of Pennsylvania, CIS 565: GPU Programming and Architecture, Project 3
- Haorong Yang
- Tested on: Windows 10 Home, i7-10750H @ 2.60GHz 16GB, GTX 2070 Super Max-Q (Personal)
This is an implementation of A Trous filter denoising based on the paper: Edge-Avoiding À-Trous Wavelet Transform for fast Global Illumination Filtering. The reference images are performed on a basic Monte-Carlo Path Tracer.
| Unsmoothed | Smoothed |
|---|---|
 |
 |
| positions | normals |
|---|---|
 |
 |
Here is a comparison of the rendering time with and without denoising. We can see that denoising always will take up extra processing time, and increases linearly with filter size.