@inproceedings{6ceddab8a1494a8fb0940d9b7f9a1338,
title = "A generalised-mutual-information-based oracle for hierarchical radiosity",
abstract = "One of the main problems in the radiosity method is how to discretise a scene into mesh elements that allow us to accurately represent illumination. In this paper we present a new refinement criterion for hierarchical radiosity based on the continuous and discrete generalised mutual information measures between two patches or elements. These measures, derived from the generalised entropy of Harvda-Charv{\'a}t-Tsallis, express the information transfer within a scene. The results obtained improve on the ones based on kernel smoothness and Shannon mutual information.",
author = "Jaume Rigau and Miquel Feixas and Mateu Sbert",
year = "2007",
doi = "10.1007/978-3-540-72586-2\_14",
language = "English",
isbn = "9783540725855",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
number = "PART 2",
pages = "105--113",
booktitle = "Computational Science - ICCS 2007 - 7th International Conference, Proceedings",
address = "Germany",
edition = "PART 2",
note = "7th International Conference on Computational Science, ICCS 2007 ; Conference date: 27-05-2007 Through 30-05-2007",
}