BEV-S: a simple and robust look-around perception network

Renfei Ryan Wong

doi:10.1117/12.3011087

20 October 2023 BEV-S: a simple and robust look-around perception network

Renfei Ryan Wong

Proceedings Volume 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023); 128141P (2023) https://doi.org/10.1117/12.3011087
Event: Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 2023, Chongqing, China

Abstract

This study explores a novel Bird’s Eye View (BEV) method for autonomous vehicles. By simplifying the converter, we have minimized computation in the process of converting 3D maps to BEV maps, leading to more effective solutions. Our understanding of the BEV model and its advantages over current solutions was informed by related works on BEV and Semantic Segmentation. Through in-depth research on similar BEV algorithms and experiments, we identified key operations crucial for the accuracy of BEV maps while eliminating unrelated operations. As a result, we developed a new algorithm model for pure electric vehicles based on a previous transformer, retaining only key operations for a more concise model. Our model may achieve higher accuracy than previous models by ensuring that critical operations are not disrupted.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Renfei Ryan Wong "BEV-S: a simple and robust look-around perception network", Proc. SPIE 12814, Third International Conference on Green Communication, Network, and Internet of Things (CNIoT 2023), 128141P (20 October 2023); https://doi.org/10.1117/12.3011087

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

Includes PDF, HTML & Video, when available