Sensor Fusion by Spatial Encoding for Autonomous Driving

TL;DR

This paper presents a novel sensor fusion method using spatial encoding with Transformers for autonomous driving, effectively combining camera and LiDAR data to improve perception accuracy in complex environments.

Contribution

Introduces a Transformer-based sensor fusion approach with multi-resolution modules for enhanced local and global contextual integration in autonomous driving.

Findings

Outperforms previous methods on challenging benchmarks

Achieves 8% and 19% higher driving scores on specific datasets

Demonstrates robustness in high-density traffic scenarios

Abstract

Sensor fusion is critical to perception systems for task domains such as autonomous driving and robotics. Recently, the Transformer integrated with CNN has demonstrated high performance in sensor fusion for various perception tasks. In this work, we introduce a method for fusing data from camera and LiDAR. By employing Transformer modules at multiple resolutions, proposed method effectively combines local and global contextual relationships. The performance of the proposed method is validated by extensive experiments with two adversarial benchmarks with lengthy routes and high-density traffics. The proposed method outperforms previous approaches with the most challenging benchmarks, achieving significantly higher driving and infraction scores. Compared with TransFuser, it achieves 8% and 19% improvement in driving scores for the Longest6 and Town05 Long benchmarks, respectively.