Fusion of neural networks, for LIDAR-based evidential road mapping

  title={Fusion of neural networks, for LIDAR-based evidential road mapping},
  author={Edouard Capellier and Franck Davoine and V{\'e}ronique Berge-Cherfaoui and You Li},
  journal={J. Field Robotics},
LIDAR sensors are usually used to provide autonomous vehicles with 3D representations of their environment. In ideal conditions, geometrical models could detect the road in LIDAR scans, at the cost of a manual tuning of numerical constraints, and a lack of flexibility. We instead propose an evidential pipeline, to accumulate road detection results obtained from neural networks. First, we introduce RoadSeg, a new convolutional architecture that is optimized for road detection in LIDAR scans… Expand


Evidential deep learning for arbitrary LIDAR object classification in the context of autonomous driving
An evidential end-to-end deep neural network is proposed to classify LIDAR objects that is capable of classifying ambiguous and incoherent objects as unknown, while only having been trained on vehicles and vulnerable road users. Expand
Evidential grid mapping, from asynchronous LIDAR scans and RGB images, for autonomous driving
An evidential fusion algorithm between LIDAR scans and RGB images that can be used to fuse contradictory information over time, while allowing the vehicle to operate even in the event of the failure of a single sensor. Expand
Hybrid conditional random field based camera-LIDAR fusion for road detection
This model integrates the information from the two sensors in a probabilistic way and makes good use of both sensors, and can be optimized efficiently with graph cuts to get road areas. Expand
Fast LIDAR-based road detection using fully convolutional neural networks
By considering a top-view representation, road detection is reduced to a single-scale problem that can be addressed with a simple and fast fully convolutional neural network (FCN), which is particularly suitable for real-time applications. Expand
Improving city model determination by using road detection from LIDAR data
A new road classification technique from LIght Detection And Ranging (LIDAR) data is presented that relies on region growing in order to classify areas as road. The new method corrects some of theExpand
SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud
An end-to-end pipeline called SqueezeSeg based on convolutional neural networks (CNN), which takes a transformed LiDAR point cloud as input and directly outputs a point-wise label map, which is then refined by a conditional random field (CRF) implemented as a recurrent layer. Expand
Road Detection Using High Resolution LIDAR
This paper proposes a road detection approach based solely on dense 3D-LIDAR data that is suitable for applications on rural areas and inner-city with unmarked roads, and obtaining promising results when compared to other state-of-art approaches. Expand
Automatic Vector-based Road Structure Mapping Using Multi-beam LiDAR
  • Xudong He, Junqiao Zhao, +4 authors Chen Ye
  • Computer Science
  • 2018 21st International Conference on Intelligent Transportation Systems (ITSC)
  • 2018
In this paper, we studied a SLAM method for vector-based road structure mapping using multi-beam LiDAR. We proposed to use the polyline as the primary mapping element instead of grid cell or pointExpand
RangeNet ++: Fast and Accurate LiDAR Semantic Segmentation
This paper proposes a novel post-processing algorithm that deals with problems arising from this intermediate representation of range images as an intermediate representation in combination with a Convolutional Neural Network exploiting the rotating LiDAR sensor model. Expand
Real-Time Road Lane Detection in Urban Areas Using LiDAR Data
This paper presents a practical real-time working prototype for road lane detection using LiDAR data, which can be further extended to automatic lane-level map generation. Expand