Related Experiment Video
Updated: Oct 28, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
725
The geometric attention-aware network for lane detection in complex road scenes
JianWu Long1, ZeRan Yan1, Lang Peng2
1College of Computer Science and Engineering, Chongqing University of Technology, Chongqing, China.
Plos One
|July 15, 2021
Summary
This study introduces a novel Geometric Attention-Aware Network (GAAN) for robust lane detection in challenging road conditions. The GAAN model demonstrates superior performance in complex scenarios, improving lane detection accuracy.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Machine Learning
Background:
- Lane detection is crucial for autonomous driving systems.
- Complex road conditions like poor lighting and irrelevant markings hinder current lane detection methods.
Purpose of the Study:
- To develop an effective lane detection model for complex road scenes.
- To improve the accuracy and robustness of lane detection algorithms.
Main Methods:
- Proposed a Geometric Attention-Aware Network (GAAN) with a multi-task branch architecture.
- Utilized an Attention Information Propagation (AIP) module for inter-branch communication.
- Employed a Geometric Attention-Aware (GAA) module for feature fusion.
Main Results:
- The GAAN model was evaluated on CULane, TuSimple, and BDD100K datasets.
- Experimental results indicate strong performance compared to existing lane detection networks.
- The proposed method effectively handles complex road scenes.
Conclusions:
- The Geometric Attention-Aware Network (GAAN) offers a promising solution for challenging lane detection tasks.
- The model's architecture facilitates robust feature fusion and accurate lane identification.
- GAAN advances the state-of-the-art in autonomous driving perception systems.
