Adaptive Multi-view Screen-space Ambient Obscurance

Abstract

Screen-space ambient obscurance techniques provide a cheap, yet visually pleasing approximation to the expensive global illumination effect. However, view-dependent artifacts frequently occur due to the limited scene information available in a single view, which forms the basis of these solutions. Recent work aims at suppressing artifacts with multiple viewpoints, but fails when occlusion occurs in the added views. We propose a novel solution that adaptively discards objects during rendering of additional cameras to minimize occlusion. Our new method provides a more robust and stable suppression of the view-dependent artifacts, and is potentially suitable for real-time rendering contexts with adequate optimization effort.