Advanced

Designing a deep learning network for self-driving vehicles

Ibrahim, Nora LU (2017) In Bachelor's Theses in Mathematical Sciences NUMK01 20171
Mathematics (Faculty of Engineering)
Abstract
Autonomous vehicles, or as we usually know them by, self-driving cars, are
often presented as the transportation technology of the future. The possibility of a vehicle to drive itself without any human intervention or manual control finally becomes possible.

The common machine learning algorithm that is being used in autonomous
vehicles is the deep learning network. Deep learning is a machine learning
technique inspired by the biological nervous system. Using deep learning
networks for classification tasks which involve curve and line detection of the roads in order to control steering angle of an autonomous vehicle has proven to be increasingly successful.

The main objective of this thesis it to create our own deep learning... (More)
Autonomous vehicles, or as we usually know them by, self-driving cars, are
often presented as the transportation technology of the future. The possibility of a vehicle to drive itself without any human intervention or manual control finally becomes possible.

The common machine learning algorithm that is being used in autonomous
vehicles is the deep learning network. Deep learning is a machine learning
technique inspired by the biological nervous system. Using deep learning
networks for classification tasks which involve curve and line detection of the roads in order to control steering angle of an autonomous vehicle has proven to be increasingly successful.

The main objective of this thesis it to create our own deep learning network
using convolutional neural networks (CNNs) in order to teach the vehicle
to drive without driver. In this thesis the CNN with small convolutional filters has been trained from scratch using our own images and a back-propagation algorithm. The resulting model is tested on a driving simulator in order to evaluate its ability to drive itself. (Less)
Popular Abstract
Autonomous vehicles, or as we usually know them by, self-driving cars, are
often presented as the transportation technology of the future. The possibility of a vehicle to drive itself without any human intervention or manual control finally becomes possible.

The common machine learning algorithm that is being used in autonomous
vehicles is the deep learning network. Deep learning is a machine learning
technique inspired by the biological nervous system. Using deep learning
networks for classification tasks which involve curve and line detection of the roads in order to control steering angle of an autonomous vehicle has proven to be increasingly successful.

The main objective of this thesis it to create our own deep learning... (More)
Autonomous vehicles, or as we usually know them by, self-driving cars, are
often presented as the transportation technology of the future. The possibility of a vehicle to drive itself without any human intervention or manual control finally becomes possible.

The common machine learning algorithm that is being used in autonomous
vehicles is the deep learning network. Deep learning is a machine learning
technique inspired by the biological nervous system. Using deep learning
networks for classification tasks which involve curve and line detection of the roads in order to control steering angle of an autonomous vehicle has proven to be increasingly successful.

The main objective of this thesis it to create our own deep learning network
using convolutional neural networks (CNNs) in order to teach the vehicle
to drive without driver. In this thesis the CNN with small convolutional filters has been trained from scratch using our own images and a back-propagation algorithm. The resulting model is tested on a driving simulator in order to evaluate its ability to drive itself. (Less)
Please use this url to cite or link to this publication:
author
Ibrahim, Nora LU
supervisor
organization
course
NUMK01 20171
year
type
M2 - Bachelor Degree
subject
keywords
Deep learning network, self-driving vehicle, autonomous, Neural network
publication/series
Bachelor's Theses in Mathematical Sciences
report number
LUNFNA-4013-2017
ISSN
1654-6229
other publication id
2017:K10
language
English
id
8918959
date added to LUP
2017-06-28 16:12:03
date last changed
2017-06-28 16:12:03
@misc{8918959,
  abstract     = {Autonomous vehicles, or as we usually know them by, self-driving cars, are
often presented as the transportation technology of the future. The possibility of a vehicle to drive itself without any human intervention or manual control finally becomes possible.

The common machine learning algorithm that is being used in autonomous
vehicles is the deep learning network. Deep learning is a machine learning
technique inspired by the biological nervous system. Using deep learning
networks for classification tasks which involve curve and line detection of the roads in order to control steering angle of an autonomous vehicle has proven to be increasingly successful.

The main objective of this thesis it to create our own deep learning network
using convolutional neural networks (CNNs) in order to teach the vehicle
to drive without driver. In this thesis the CNN with small convolutional filters has been trained from scratch using our own images and a back-propagation algorithm. The resulting model is tested on a driving simulator in order to evaluate its ability to drive itself.},
  author       = {Ibrahim, Nora},
  issn         = {1654-6229},
  keyword      = {Deep learning network,self-driving vehicle,autonomous,Neural network},
  language     = {eng},
  note         = {Student Paper},
  series       = {Bachelor's Theses in Mathematical Sciences},
  title        = {Designing a deep learning network for self-driving vehicles},
  year         = {2017},
}