Self-organized Learning of 3 Dimensions. Szepesvári, C. and Lörincz, A. In Marinaro, M. and Morasso, P., editors, ICANN, volume 2, pages 671–674, Sorrento, Italy, 1994. IEEE.
Paper abstract bibtex The geometric learning capabilities of a competitive neural network are studied. It is shown that the appropriate selection of a neural activity function enables the learning of the 3D geometry of a world, from two of the 2D projections of 3D extended objects
@inproceedings{szepesvari1994a,
abstract = {The geometric learning capabilities of a competitive neural network are studied. It is shown that the appropriate selection of a neural activity function enables the learning of the 3D geometry of a world, from two of the 2D projections of 3D extended objects},
address = {Sorrento, Italy},
author = {Szepesv{\'a}ri, Cs. and L{\"o}rincz, A.},
booktitle = {ICANN},
date-added = {2010-08-28 17:38:14 -0600},
date-modified = {2010-11-25 00:58:22 -0700},
editor = {Marinaro, M. and Morasso, P.G.},
keywords = {manifold learning, theory, neural networks},
pages = {671--674},
publisher = {IEEE},
title = {Self-organized Learning of 3 Dimensions},
url_paper = {icann94.ps.pdf},
volume = {2},
year = {1994}}