In many machine learning applications, one has not only training data but also some high-level information about certain invariances that the system should exhibit. In character recognition, for example, the answer should be invariant with respect to small spatial distortions in the input images (translations, rotations, scale changes, etcetera). We have implemented a scheme that minimizes the derivative of the classifier outputs with respect to distortion operators of our choosing. This not only produces tremendous speed advantages, but also provides a powerful language for specifying what generalizations we wish the network to perform.
|Title of host publication
|Subtitle of host publication
|Pattern Recognition Methodology and Systems
|Institute of Electrical and Electronics Engineers Inc.
|Number of pages
|Published - 1992
|11th IAPR International Conference on Pattern Recognition, IAPR 1992 - The Hague, Netherlands
Duration: Aug 30 1992 → Sep 3 1992
|Proceedings - International Conference on Pattern Recognition
|11th IAPR International Conference on Pattern Recognition, IAPR 1992
|8/30/92 → 9/3/92
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition