Experiments with an extended tangent distance
- 1 January 2000
- proceedings article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 38-42
- https://doi.org/10.1109/icpr.2000.906014
Abstract
Invariance is an important aspect in image object recog- nition. We present results obtained with an extended tangen t distance incorporated in a kernel density based Bayesian classifier to compensate for affine image variations. An im- age distortion model for local variations is introduced and its relationship to tangent distance is considered. The pro - posed classification algorithms are evaluated on databases of different domains. An excellent result of 2.2% error rate on the original USPS handwritten digits recognition task is obtained. On a database of radiographs from daily routine, best results are obtained by combining tangent distance and the proposed distortion model.Keywords
This publication has 5 references indexed in Scilit:
- Invariant image object recognition using mixture densitiesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2000
- Transformation Invariance in Pattern Recognition — Tangent Distance and Tangent PropagationPublished by Springer Nature ,1998
- Invariant pattern recognition: A reviewPattern Recognition, 1996
- Combining classifiersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1996
- A Bayesian similarity measure for direct image matchingPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1996