Etienne Barnard and David Casasent,
Invariance and Neural Nets,
IEEE Transactions on Neural Networks,
2, 5, pp. 498-508, September 1991.
Etienne Barnard and David Casasent,
Invariance and Neural Nets,
IEEE Transactions on Neural Networks,
2, 5, pp. 498-508, September 1991.Invariance with respect to certain transformations is one of the main tasks of pattern recognition systems. We study various techniques for obtaining this invariance with neural net classifiers and identify the invariant-feature techniques as the most suitable for current neural classifiers. A new formulation in terms of constraints on the feature values leads to a general method for transforming any given feature space so that it becomes invariant to specified transformations. A case study using range imagery is used to exemplify these ideas, and good performance is obtained.