Prediction of steering wheel angle at night based on CNN

Zhuolun Li

doi:10.1117/12.2673094

14 April 2023 Prediction of steering wheel angle at night based on CNN

Zhuolun Li

Author Affiliations +

Proceedings Volume 12612, International Conference on Artificial Intelligence and Industrial Design (AIID 2022); 126120A (2023) https://doi.org/10.1117/12.2673094
Event: International Conference on Artificial Intelligence and Industrial Design (AIID 2022), 2022, Zhuhai, China

Abstract

In the past few years, the development of autopilot has made great progress. Autonomous vehicle is composed of many parts, including environmental perception and decision control. Steering angle prediction is particularly important in the control of autonomous vehicle and has attracted extensive attention. However, most of the existing models study the behavior methods during the day and have poor compatibility to night traffic environment. This paper is conducted to overcome this shortcoming through deep learning, which is supposed to automatically study relevant features from driving data without manual intervention. Specifically, the convolutional neural network (CNN) with residual structure has been used to build a model for learning and training, and the training data comes from the unity simulator. Based on this model, the steering wheel angle is predicted. The prediction results show that compared with other learning methods based on CNN, this model has higher accuracy in the case of poor lighting at night, and could be able to conform to several driving scenes as well as less revisions.

Citation Download Citation

Zhuolun Li "Prediction of steering wheel angle at night based on CNN", Proc. SPIE 12612, International Conference on Artificial Intelligence and Industrial Design (AIID 2022), 126120A (14 April 2023); https://doi.org/10.1117/12.2673094

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE