Positional-encoding-based general obstacle detection method for autonomous driving

Chuanxiang Lin

doi:10.1117/12.3051605

20 January 2025 Positional-encoding-based general obstacle detection method for autonomous driving

Chuanxiang Lin

Author Affiliations +

Proceedings Volume 13422, Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024); 1342214 (2025) https://doi.org/10.1117/12.3051605
Event: Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024), 2024, Xi'an, China

Abstract

To address the limitations of traditional whitelist-based obstacle detection systems in handling complex scenarios for autonomous driving, we propose a novel approach leveraging PETR (Position Embedding Transformation Representation). Our method predicts three-dimensional (3D) drivable areas by modeling occupancy information, employing a new 3D occupancy query mechanism that integrates 2D image features through an attention decoder. Extensive ablation studies identify optimal backbone networks and query sizes, demonstrating the method's ability to accurately model driving scenarios and delineate drivable areas. This innovation significantly enhances general obstacle detection in autonomous driving.

(2025) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Chuanxiang Lin "Positional-encoding-based general obstacle detection method for autonomous driving", Proc. SPIE 13422, Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024), 1342214 (20 January 2025); https://doi.org/10.1117/12.3051605

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