RGB and LiDAR fusion based 3D Semantic Segmentation for Autonomous Driving

TL;DR

This paper proposes sensor fusion architectures combining RGB images and LiDAR data for 3D semantic segmentation in autonomous driving, achieving significant accuracy improvements on the KITTI dataset.

Contribution

It introduces novel early, mid-level, and hybrid fusion architectures that convert RGB images into a polar-grid format for improved 3D segmentation.

Findings

Achieved 10% mIoU improvement over LiDAR-only baselines.

Enhanced segmentation accuracy for cars, pedestrians, and cyclists.

Validated on KITTI dataset with two state-of-the-art models.

Abstract

LiDAR has become a standard sensor for autonomous driving applications as they provide highly precise 3D point clouds. LiDAR is also robust for low-light scenarios at night-time or due to shadows where the performance of cameras is degraded. LiDAR perception is gradually becoming mature for algorithms including object detection and SLAM. However, semantic segmentation algorithm remains to be relatively less explored. Motivated by the fact that semantic segmentation is a mature algorithm on image data, we explore sensor fusion based 3D segmentation. Our main contribution is to convert the RGB image to a polar-grid mapping representation used for LiDAR and design early and mid-level fusion architectures. Additionally, we design a hybrid fusion architecture that combines both fusion algorithms. We evaluate our algorithm on KITTI dataset which provides segmentation annotation for cars, pedestrians and cyclists. We evaluate two state-of-the-art architectures namely SqueezeSeg and PointSeg and improve the mIoU score by 10 % in both cases relative to the LiDAR only baseline.