END TO END LEARNING FOR A DRIVING SIMULATOR

Autor: V. F. Alexeev, A. I. Staravoitau, G. A. Piskun, D. V. Likhacheuski
Jazyk: ruština
Rok vydání: 2019
Předmět:
Zdroj: Doklady Belorusskogo gosudarstvennogo universiteta informatiki i radioèlektroniki, Vol 0, Iss 2, Pp 85-91 (2019)
Druh dokumentu: article
ISSN: 1729-7648
Popis: Convolutional network approach is utilized for training an end-to-end model that would let a car drive itself around the track in a driving simulator by predicting steering angles based on the simulated camera data.
Databáze: Directory of Open Access Journals