Efficient Pattern Recognition Using a New Transformation Distance

By Patrice Y. Simard, Yann LeCun, John S. Denker. 1992

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

A Geometric Analysis of Subspace Clustering with Outliers (2011, background): “[36] characterize this invariance with a 7-dimensional manifold; that is, different transformations of a single digit are well approximated by a 7dimensional manifold.”
Distance Metric Learning for Large Margin Nearest Neighbor Classification (2005, background): “Nevertheless, it often yields competitive results, and in certain domains, when cleverly combined with prior knowledge, it has significantly advanced the state-ofthe-art [1] [14].”
A Simple Geometric Interpretation of SVM using Stochastic Adversaries (2012, methodology): “For example, tangentdistances [23] have been used to incorporate invariances in the training data.”
Learning from one example through shared densities on transforms (2000, methodology): “The method of creating artificial data discussed above is similar to the methods used in [11] and [12] in which extra data is created by randomly sampling from affine transforms and applying those transforms to data samples already observed.”
SVM-KNN: Discriminative Nearest Neighbor Classification for Visual Category Recognition (2006, background): “Examples are tangent distance on the USPS zip code dataset (Simard, LeCun & Denker [35]), shape context based distance on the MNIST digit dataset (Belongie, Malik & Puzicha [1]), distances between histograms of textons on the CUReT data set (Leung and Malik [22], Varma and Zisserman [40]), and geome”
Content-based Image Retrieval in Medical Applications (2004, methodology): “The feature values obtained for each image are combined to a feature vector, which is then used for k-nearest-neighbor classification based on e.g. the Euclidean, or Simard’s tangent distance [20].”
Distance Metric Learning for Large Margin Nearest Neighbor Classification (2006, background): “Despite its simplicity, the kNN rule often yields competitive results and in certain domains, when cleverly combined with prior knowledge, it has significantly advanced the state-of-the-art (Belongie et al., 2002; Simard et al., 1993).”
Regularization Networks and Support Vector Machines (2000, background): “[62] showed how several invariances can be embedded in the stabilizer or, equivalently, in virtual examples (see for a related work on tangent distance [89] and [84]).”
Experiments with a New Boosting Algorithm (1996, background): “Using the more sophisticated tangent distance [21] is in our future plans.”
The Manifold Tangent Classifier (2011, methodology): “The tangent distance (Simard et al., 1993) and TangentProp (Simard et al.”