Self-driving a Car in Simulation Through a CNN

Javier del Egio; Luis Miguel Bergasa; Eduardo Romera; Carlos Gómez Huélamo; Javier Araluce; Rafael Barea

doi:10.1007/978-3-319-99885-5_3

Self-driving a Car in Simulation Through a CNN

Javier del Egio^*
, Luis Miguel Bergasa
, Eduardo Romera
, Carlos Gómez Huélamo
, Javier Araluce
, Rafael Barea

^*Corresponding author for this work

University of Alcalá

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

This work presents a comparison between different Convolutional Neural Network models, testing its performance when it leads a self-driving car in a simulated environment. To do so, driving data has been obtained manually driving the simulator as ground truth and different network models with diverse complexity levels has been created and trained with the data previously obtained using end-to-end deep learning techniques. Once this CNNs are trained, they are tested in the driving simulator, checking their ability of minimizing the car distance to the center of the lane, its heading error and its RMSE. The neural networks will be evaluated according to these parameters. Finally, conclusions will be drawn about the performance of the different models according to the parameters mentioned before in order to find the optimum CNN for the developed application.

Original language	English
Title of host publication	Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018
Editors	Raquel Fuentetaja Pizán, Ángel García Olaya, Maria Paz Sesmero Lorente, Jose Antonio Iglesias Martínez, Agapito Ledezma Espino
Publisher	Springer Verlag
Pages	31-43
Number of pages	13
ISBN (Print)	9783319998848
DOIs	https://doi.org/10.1007/978-3-319-99885-5_3
Publication status	Published - 2019
Externally published	Yes
Event	19th International Conference Workshop of Physical Agents, WAF 2018 - Madrid, Spain Duration: 22 Nov 2018 → 23 Nov 2018

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	855
ISSN (Print)	2194-5357

Conference

Conference	19th International Conference Workshop of Physical Agents, WAF 2018
Country/Territory	Spain
City	Madrid
Period	22/11/18 → 23/11/18

Keywords

Convolutional Neural Network (CNN)
Self-driving

Access to Document

10.1007/978-3-319-99885-5_3

Cite this

del Egio, J., Bergasa, L. M., Romera, E., Gómez Huélamo, C., Araluce, J., & Barea, R. (2019). Self-driving a Car in Simulation Through a CNN. In R. F. Pizán, Á. G. Olaya, M. P. S. Lorente, J. A. I. Martínez, & A. L. Espino (Eds.), Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018 (pp. 31-43). (Advances in Intelligent Systems and Computing; Vol. 855). Springer Verlag. https://doi.org/10.1007/978-3-319-99885-5_3

del Egio, Javier ; Bergasa, Luis Miguel ; Romera, Eduardo et al. / Self-driving a Car in Simulation Through a CNN. Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018. editor / Raquel Fuentetaja Pizán ; Ángel García Olaya ; Maria Paz Sesmero Lorente ; Jose Antonio Iglesias Martínez ; Agapito Ledezma Espino. Springer Verlag, 2019. pp. 31-43 (Advances in Intelligent Systems and Computing).

@inproceedings{cffc1240256a42c6ae83128009bde564,

title = "Self-driving a Car in Simulation Through a CNN",

abstract = "This work presents a comparison between different Convolutional Neural Network models, testing its performance when it leads a self-driving car in a simulated environment. To do so, driving data has been obtained manually driving the simulator as ground truth and different network models with diverse complexity levels has been created and trained with the data previously obtained using end-to-end deep learning techniques. Once this CNNs are trained, they are tested in the driving simulator, checking their ability of minimizing the car distance to the center of the lane, its heading error and its RMSE. The neural networks will be evaluated according to these parameters. Finally, conclusions will be drawn about the performance of the different models according to the parameters mentioned before in order to find the optimum CNN for the developed application.",

keywords = "Convolutional Neural Network (CNN), Self-driving",

author = "\{del Egio\}, Javier and Bergasa, \{Luis Miguel\} and Eduardo Romera and \{G{\'o}mez Hu{\'e}lamo\}, Carlos and Javier Araluce and Rafael Barea",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 19th International Conference Workshop of Physical Agents, WAF 2018 ; Conference date: 22-11-2018 Through 23-11-2018",

year = "2019",

doi = "10.1007/978-3-319-99885-5\_3",

language = "English",

isbn = "9783319998848",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer Verlag",

pages = "31--43",

editor = "Piz{\'a}n, \{Raquel Fuentetaja\} and Olaya, \{{\'A}ngel Garc{\'i}a\} and Lorente, \{Maria Paz Sesmero\} and Mart{\'i}nez, \{Jose Antonio Iglesias\} and Espino, \{Agapito Ledezma\}",

booktitle = "Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018",

address = "Germany",

}

del Egio, J, Bergasa, LM, Romera, E, Gómez Huélamo, C, Araluce, J & Barea, R 2019, Self-driving a Car in Simulation Through a CNN. in RF Pizán, ÁG Olaya, MPS Lorente, JAI Martínez & AL Espino (eds), Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018. Advances in Intelligent Systems and Computing, vol. 855, Springer Verlag, pp. 31-43, 19th International Conference Workshop of Physical Agents, WAF 2018, Madrid, Spain, 22/11/18. https://doi.org/10.1007/978-3-319-99885-5_3

Self-driving a Car in Simulation Through a CNN. / del Egio, Javier; Bergasa, Luis Miguel; Romera, Eduardo et al.
Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018. ed. / Raquel Fuentetaja Pizán; Ángel García Olaya; Maria Paz Sesmero Lorente; Jose Antonio Iglesias Martínez; Agapito Ledezma Espino. Springer Verlag, 2019. p. 31-43 (Advances in Intelligent Systems and Computing; Vol. 855).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Self-driving a Car in Simulation Through a CNN

AU - del Egio, Javier

AU - Bergasa, Luis Miguel

AU - Romera, Eduardo

AU - Gómez Huélamo, Carlos

AU - Araluce, Javier

AU - Barea, Rafael

PY - 2019

Y1 - 2019

N2 - This work presents a comparison between different Convolutional Neural Network models, testing its performance when it leads a self-driving car in a simulated environment. To do so, driving data has been obtained manually driving the simulator as ground truth and different network models with diverse complexity levels has been created and trained with the data previously obtained using end-to-end deep learning techniques. Once this CNNs are trained, they are tested in the driving simulator, checking their ability of minimizing the car distance to the center of the lane, its heading error and its RMSE. The neural networks will be evaluated according to these parameters. Finally, conclusions will be drawn about the performance of the different models according to the parameters mentioned before in order to find the optimum CNN for the developed application.

AB - This work presents a comparison between different Convolutional Neural Network models, testing its performance when it leads a self-driving car in a simulated environment. To do so, driving data has been obtained manually driving the simulator as ground truth and different network models with diverse complexity levels has been created and trained with the data previously obtained using end-to-end deep learning techniques. Once this CNNs are trained, they are tested in the driving simulator, checking their ability of minimizing the car distance to the center of the lane, its heading error and its RMSE. The neural networks will be evaluated according to these parameters. Finally, conclusions will be drawn about the performance of the different models according to the parameters mentioned before in order to find the optimum CNN for the developed application.

KW - Convolutional Neural Network (CNN)

KW - Self-driving

UR - https://www.scopus.com/pages/publications/85057779992

U2 - 10.1007/978-3-319-99885-5_3

DO - 10.1007/978-3-319-99885-5_3

M3 - Conference contribution

AN - SCOPUS:85057779992

SN - 9783319998848

T3 - Advances in Intelligent Systems and Computing

SP - 31

EP - 43

BT - Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018

A2 - Pizán, Raquel Fuentetaja

A2 - Olaya, Ángel García

A2 - Lorente, Maria Paz Sesmero

A2 - Martínez, Jose Antonio Iglesias

A2 - Espino, Agapito Ledezma

PB - Springer Verlag

T2 - 19th International Conference Workshop of Physical Agents, WAF 2018

Y2 - 22 November 2018 through 23 November 2018

ER -

del Egio J, Bergasa LM, Romera E, Gómez Huélamo C, Araluce J, Barea R. Self-driving a Car in Simulation Through a CNN. In Pizán RF, Olaya ÁG, Lorente MPS, Martínez JAI, Espino AL, editors, Advances in Physical Agents - Proceedings of the 19th International Workshop of Physical Agents WAF 2018. Springer Verlag. 2019. p. 31-43. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-319-99885-5_3

Self-driving a Car in Simulation Through a CNN

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this