Classification of GPU rendering errors with Artificial Neural Networks
Alexander Hansson, Peter Moodie · 2019
Image quality metrics are used to evaluate the percieved quality of processed images. Differences in hardware between graphics processors contribute to noise during quality evaluation. In this masters thesis paper we train and evaluate neural networks as metrics to evaluate GPU rendering quality. The neural networks can successfully ignore the rendering noise that occurs when the test and reference frames are rendered by different GPUs. This reduces tedious human interaction which requires manual updates of reference-frames during quality testing.