Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark for Bird-Eye View Mapping in Autonomous Driving: Dur360BEV_Dataset_split_02 [dataset]

Breckon, Toby P. (Durham University, UK); Atapour-Abarghouei, Amir (Durham University, UK); A. Gaus, Yona Falinie (Durham University, UK); Wenke, E (Durham University, UK); Sun, Yixin (Durham University, UK); Li, Li (Durham University, UK)

Actions

Download Analytics Citations

Export to: EndNote | Zotero | Mendeley

Collections

This file is in the following collections:

Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark for Bird-Eye View Mapping in Autonomous Driving [dataset]

Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark for Bird-Eye View Mapping in Autonomous Driving: Dur360BEV_Dataset_split_02 [dataset] Open Access

We present Dur360BEV, a novel single spherical camera autonomous driving dataset equipped with a high-resolution 128 channel 3D LiDAR and RTK-refined GNSS/INS localisation, as well as a sample benchmark architecture designed to generate Bird's-Eye-View (BEV) maps using only the single spherical camera input. This dataset and benchmark address the challenges of BEV generation in autonomous driving, particularly by reducing hardware complexity through the use of a single 360-degree camera instead of multiple perspective view cameras. Within our benchmark architecture, we propose a novel spherical-image-to-BEV module that leverages spherical imagery and a refined sampling strategy to project features from 2D to 3D. Our approach also includes an innovative application of focal loss, specifically adapted to address the extreme class imbalance often encountered in BEV segmentation tasks, that demonstrates improved segmentation performance on the Dur360BEV dataset. The results show that our benchmark not only simplifies the sensor setup but also achieves competitive performance against contemporary state of the art approaches.

Descriptions

Resource type: Dataset
Contributors: Creator: Wenke, E ¹
Editor: Wenke, E ¹
Creator: Breckon, Toby P. ¹
Editor: Breckon, Toby P. ¹
Editor: Li, Li ¹
Editor: Sun, Yixin ¹
Editor: A. Gaus, Yona Falinie ¹
Editor: Atapour-Abarghouei, Amir ¹

¹ Durham University, UK
Funder: Durham University
Research methods: Data collection performed in the UK, under Durham University Ethics Approval Ref: COMP-2021-10-03T23-38-21-qhww73
Other description
Keyword: high resolution LiDAR
dataset
autonomous driving
flash LiDAR
spherical camera
360 degree camera
bird's eye view
BEV
Subject: Computer Science
Artificial intelligence
Autonomy
Computer vision--Data processing
Location: Durham, England, United Kingdom
Language: English
Cited in: arxiv:2503.00675
Identifier: ark:/32150/r3zp38wc60z
Rights: Creative Commons Attribution 4.0 International (CC BY)
Publisher: Durham University
Date Created: March 2024

File Details

Depositor: T. Breckon
Date Uploaded: 11 March 2025, 14:03:54
Date Modified: 26 March 2025, 14:03:21
Audit Status: Audits have not yet been run on this file.
Characterization: not yet characterized

Activity of users you follow
User Activity	Date
User N. Syrotiuk has updated Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark for Bird-Eye View Mapping in Autonomous Driving: Dur360BEV_Dataset_split_02 [dataset]	7 days ago
User W. E has updated Dur360BEV: A Real-world 360-degree Single Camera Dataset and Benchmark for Bird-Eye View Mapping in Autonomous Driving: Dur360BEV_Dataset_split_02 [dataset]	13 days ago
User N. Syrotiuk has updated Dur360BEV_Dataset_split_02	22 days ago
User N. Syrotiuk has updated Dur360BEV_Dataset_split_02	22 days ago