"Multi-View LiDAR Dataset for Pedestrian Observability in Autonomous Driving"

Read full text Summarise with VIRAT
Authors: Juan Ortiz
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

"This dataset provides multi-view LiDAR recordings of pedestrians collected in real-world conditions using a dual-sensor configuration, including a roof-mounted Top LiDAR and a Front-Right LiDAR. The dataset is specifically designed to support the analysis of three-dimensional pedestrian observability as a function of sensor placement and viewpoint. It includes synchronized point clouds acquired at varying distances (1\u201320 m) and perspectives, enabling a detailed evaluation of geometric completeness, volumetric representation, and spatial distribution of pedestrian data. By capturing complementary viewpoints, the dataset allows researchers to investigate multi-view consistency and the impact of sensor configuration on perception quality. Additionally, the provided metadata facilitates temporal alignment and reproducible analysis. This resource supports research in sensor configuration, multi-view perception, and geometry-aware evaluation, contributing to the development of more robust perception systems in autonomous driving applications."

Fields of Study & Subjects

Autonomous Vehicle Technology and Safety Advanced Optical Sensing Technologies Robotics and Sensor-Based Localization

Publication details

Journal / source
IEEE DataPort
Publication date
25 March 2026
Language
EN
Document type
Dataset