Título: Experimentos em Aprendizado de Máquina para Fusão de Sensores
Autores: Faceli, Katti; Carvalho, André C. P. L. F. de; Rezende, Solange O.
Resumo: Robot navigation systems need accurate distance measures to describe the robot environment (distances to target and obstacles). These measures are acquired by sensors, which usually do not have the necessary precision. In order to obtain more accurate distance measures, this paper investigates the behaviour of machine learning techniques for sensor fusion. For such, MultiLayer Perceptron networks and Support Vector Machines are used. The accuracy of the distances measured by the fusion models are compared to those achieved by the individual sensors.
Código DOI: 10.21528/CBRN2001-068
Artigo em pdf: 5cbrn_068.pdf
Arquivo BibTex: 5cbrn_068.bib