An end-to-end approach to autonomous vehicle control using deep learning
Carregando...
Data
2021
Data de publicaĆ§Ć£o:
Autores IPEN
Orientador
TĆtulo da Revista
ISSN da Revista
TĆtulo do Volume
Ć parte de
Ć parte de
Ć parte de
Revista Brasileira de ComputaĆ§Ć£o Aplicada
Resumo
The objective of this work is to develop an autonomous vehicle controller inside Grand Theft Auto V game, used as a
simulation environment. It is used an end-to-end approach, in which the model maps directly the inputs from the
image of a car hood camera and a sequence of speed values to three driving commands: steering wheel angle, accelerator
pedal pressure and brake pedal pressure. The developedmodel is composed of a convolutional neural network and a
recurring neural network. The convolutional network processes the images and the recurrent network processes the
speed data. Themodel learns fromdata generated by a human driverĀ“s commands. Two interfaces are developed: one
for collecting in-game training data and another to verify the performance of themodel for the autonomous vehicle
control. The results show that themodel after training is capable to drive the vehicle as well as a human driver. This
proves that a combination of a convolutional network with a recurrent network, using an end-to-end approach, is
capable of obtaining a good driving performance even using only images and speed velocity as sensory data.
Como referenciar
NOVELLO, GUSTAVO A.M.; YAMAMOTO, HENRIQUE Y.; CABRAL, EDUARDO L.L. An end-to-end approach to autonomous vehicle control using deep learning. Revista Brasileira de ComputaĆ§Ć£o Aplicada, v. 13, n. 3, p. 32-41, 2021. DOI: 10.5335/rbca.v13i3.12135. DisponĆvel em: http://repositorio.ipen.br/handle/123456789/32792. Acesso em: 28 Mar 2024.
Esta referĆŖncia Ć© gerada automaticamente de acordo com as normas do estilo IPEN/SP (ABNT NBR 6023) e recomenda-se uma verificaĆ§Ć£o final e ajustes caso necessĆ”rio.