Input Reconstruction Reliability Estimation - Robotics Institute Carnegie Mellon University

Input Reconstruction Reliability Estimation

Dean Pomerleau
Conference Paper, Proceedings of (NeurIPS) Neural Information Processing Systems, pp. 279 - 286, November, 1992

Abstract

The ability to identify and reason about novel aspects of their input would greatly enhance the capabilities of artificial neural networks. The extent of the novelty could be used to judge the appropriateness of individual networks for the task to be performed. The location and shape of novel features could be employed to identify the unusual components of the input and to choose an appropriate response.

BibTeX

@conference{Pomerleau-1992-15920,
author = {Dean Pomerleau},
title = {Input Reconstruction Reliability Estimation},
booktitle = {Proceedings of (NeurIPS) Neural Information Processing Systems},
year = {1992},
month = {November},
pages = {279 - 286},
}