Classification of trajectories—Extracting invariants with a neural network

Margit Kinder, Wilfried Brauer

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

A neural classifier of planar trajectories is presented. There already exist a large variety of classifiers that are specialized in particular invariants contained in a trajectory classification task such as position-invariance, rotation-invariance, and size-invariance. That is, there exist classifiers specialized in recognizing trajectories, e.g., independently of their position. The neural classifier presented in this paper is not restricted to certain invariants in a task: The neural network itself extracts the invariants contained in a classification task by assessing only the trajectories. The trajectories need to be given as a set of points. No additional information must be available for training, which saves the designer from determining the needed invariants by himself. Besides its applicability to real-world problems, such a more general classifier is also cognitively plausible: In assessing trajectories for classification, human beings are able to find class specific features no matter what kinds of invariants they are confronted with. Invariants are easily handled by ignoring unspecific features.

Original languageEnglish
Pages (from-to)1011-1017
Number of pages7
JournalNeural Networks
Volume6
Issue number7
DOIs
StatePublished - 1993

Keywords

  • Delta rule
  • Distributed encoding
  • Invariant representation
  • Perceptron
  • Task-dependent similarity measure
  • Topological
  • Tuple coding
  • Wickelfeatures

Fingerprint

Dive into the research topics of 'Classification of trajectories—Extracting invariants with a neural network'. Together they form a unique fingerprint.

Cite this