Multi-scale detail enhancement of Fattal et al. [2007] (left) compared to results produced with our decompositions (right). We are able to achieve more aggressive detail enhancement and exaggeration, without introducing artifacts.
 |
Here, we demonstrate that we can generate highly exaggerated detail even from a single input image (right), rather than three multi-light images used by Fattal et al. [2007] (left).
 |
 |