SNE-RoadSeg: Incorporating Surface Normal Information into Semantic Segmentation for Accurate Freespace Detection

TL;DR

This paper introduces SNE-RoadSeg, a novel CNN architecture that fuses RGB images with surface normal information inferred from depth data to improve freespace detection accuracy in autonomous driving, validated on a new large-scale synthetic dataset.

Contribution

The paper presents a new surface normal estimator module and a fusion-based CNN architecture for enhanced freespace detection, along with a large synthetic dataset for research.

Findings

SNE improves the performance of existing CNNs in freespace detection.

SNE-RoadSeg achieves superior accuracy across multiple datasets.

The large-scale R2D dataset supports robust model training and evaluation.

Abstract

Freespace detection is an essential component of visual perception for self-driving cars. The recent efforts made in data-fusion convolutional neural networks (CNNs) have significantly improved semantic driving scene segmentation. Freespace can be hypothesized as a ground plane, on which the points have similar surface normals. Hence, in this paper, we first introduce a novel module, named surface normal estimator (SNE), which can infer surface normal information from dense depth/disparity images with high accuracy and efficiency. Furthermore, we propose a data-fusion CNN architecture, referred to as RoadSeg, which can extract and fuse features from both RGB images and the inferred surface normal information for accurate freespace detection. For research purposes, we publish a large-scale synthetic freespace detection dataset, named Ready-to-Drive (R2D) road dataset, collected under different illumination and weather conditions. The experimental results demonstrate that our proposed SNE module can benefit all the state-of-the-art CNNs for freespace detection, and our SNE-RoadSeg achieves the best overall performance among different datasets.