A model for probabilistic independent component subspace analysis is developed and applied to texture description. Experiments show it to perform comparably to a Gaussian model, and to be useful mainly for problems in which the detection of little occurring, highfrequency image elements is important.
CITATION STYLE
de Ridder, D., Duin, R. P. W., & Kittler, J. (2002). Texture description by independent components. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2396, pp. 587–596). Springer Verlag. https://doi.org/10.1007/3-540-70659-3_61
Mendeley helps you to discover research relevant for your work.