Options
Learning a deep neural net policy for end-to-end control of autonomous vehicles
Publikationstyp
Conference Paper
Date Issued
2017
Sprache
English
Institut
TORE-URI
Start Page
4914
End Page
4919
Citation
Proceedings of the American Control Conference: 4914-4919 (2017)
Contribution to Conference
Publisher DOI
Scopus ID
ISBN of container
978-1-5090-5992-8
Deep neural networks are frequently used for computer vision, speech recognition and text processing. The reason is their ability to regress highly nonlinear functions. We present an end-to-end controller for steering autonomous vehicles based on a convolutional neural network (CNN). The deployed framework does not require explicit hand-engineered algorithms for lane detection, object detection or path planning. The trained neural net directly maps pixel data from a front-facing camera to steering commands and does not require any other sensors. We compare the controller performance with the steering behavior of a human driver.
DDC Class
600: Technik
More Funding Information
Supported by the German Academic Exchange Service (DAAD).