Efficient Training of Artificial Neural Networks for Autonomous Navigation

Neural ComputationPublished 1 February 1991

Dean Pomerleau

DOI10.1162/neco.1991.3.1.88

Source URL

Citations693

SJR quartileQ1

SJR score0.83

SNIP1.45

TL;DR

The training techniques that allow ALVINN to learn in under 5 minutes to autonomously control the Navlab by watching the reactions of a human driver are described.

Abstract

The ALVINN (Autonomous Land Vehicle In a Neural Network) project addresses the problem of training artificial neural networks in real time to perform difficult perception tasks. ALVINN is a backpropagation network designed to drive the CMU Navlab, a modified Chevy van. This paper describes the training techniques that allow ALVINN to learn in under 5 minutes to autonomously control the Navlab by watching the reactions of a human driver. Using these techniques, ALVINN has been trained to drive in a variety of circumstances including single-lane paved and unpaved roads, and multilane lined and unlined roads, at speeds of up to 20 miles per hour.

Keywords

Computer ScienceEngineering

Neural ComputationBackpropagation Applied to Handwritten Zip Code Recognition

11,878 Citations1989Yann LeCun, Bernhard E. Boser +5 more

This paper demonstrates how constraints from the task domain can be integrated into a backpropagation network through the architecture of the network, successfully applied to the recognition of handwritten zip code digits provided by the U.S. Postal Service.

Research Showcase @ Carnegie Mellon University (Carnegie Mellon University)ALVINN, an autonomous land vehicle in a neural network

1,430 Citations2018Dean Pomerleau

View PDF

IEEE Transactions on Pattern Analysis and Machine IntelligenceVision and navigation for the Carnegie-Mellon Navlab

771 Citations1988C. Thorpe, Martial Hebert +2 more

IEEE International Conference on Neural NetworksNeural network simulation at Warp speed: how we got 17 million connections per second

161 Citations1988Pomerleau, Gusciora +2 more

Results indicate that linear systolic array machines can be efficient neural network simulators and are about eight times faster at simulating the NETtalk text-to-speech network than the fastest back-propagation simulator previously reported in the literature.

Phoneme Recognition: Neural Networks vs. Hidden Markov Models

129 Citations1988Alex Waibel, Toshiyuki Hanazawa +3 more

View PDF

IFAC Proceedings VolumesAutonomous High Speed Road Vehicle Guidance by Computer Vision 1

93 Citations1987Ernst D. Dickmanns, A. Zapp

A visual feedback control system which is able to guide road vehicle on well structured roads at high speeds using contour correlation and curvature models together with the laws of perspective projection is described.

Kluwer international series in engineering and computer scienceNeural Network Based Autonomous Navigation

79 Citations1990Dean Pomerleau

This research presents a new generation of autonomous navigation systems based on traditional image processing and pattern recognition techniques that perform well under certain conditions but have problems with others.

Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIEColor Vision For Road Following

67 Citations1989Jill D. Crisman, C. Thorpe

The first system, SCARF (Supervised Classification Applied to Road Following), is designed to be fast and robust when the vehicle is running in both sunshine and shadows under constant illumination, and UNSCARF is slower, but provides good results even if the sun is alternately covered by clouds or uncovered.

Kluwer international series in engineering and computer scienceExplicit Models for Robot Road Following

42 Citations1990Karl C. Kluge, C. Thorpe

Phoneme Recognition: Neural Networks vs

26 Citations1988Alexander Waibel, Hanazawa G. Hinton +1 more

It is shown that the TDNN "invented" well-known acoustic-phonetic features and the temporal relationships between them are indeendent of position in time and hence not blurred by temporal shifts in the input.