LPLD - Learning to Predict Localized Distortions in Rendered Images

2013 - 2017, Martin Cadik ([email protected], http://cadik.posvete.cz/)

developers: Martin Cadik, Robert Herzog, Peter Harman, Lukas Teuer

In this work, we present a new data-driven full-reference image quality metric which outperforms current state-of-the-art metrics. The metric was trained on subjective ground truth data combining two publicly available datasets. For the sake of completeness we create a new testing synthetic dataset including experimentally measured subjective distortion maps. Finally, using the same machine-learning framework we optimize the parameters of popular existing metrics.

If you find LPLD project useuful, please acknowledge it by citing the following paper:

@article{ cadik13learning, 
  author    = {Martin {\v{C}}ad\'{i}k and
               Robert Herzog and
               Rafal Mantiuk and
               Radoslaw Mantiuk and
               Karol Myszkowski and
               Hans{-}Peter Seidel},
  title     = {Learning to Predict Localized Distortions in Rendered Images},
  journal   = {Comput. Graph. Forum},
  year      = {2013},
  volume    = {32},
  number    = {7},
  pages     = {401--410},
  doi       = {10.1111/cgf.12248},
}

Additional Materials

[Paper (pdf)] [Supplementary Material (html)] [Presentation slides (pdf)]