Semi-automatic high-accuracy labelling tool for multi-modal long-range sensor dataset
Authors
Izquierdo Gonzalo, Rubén; Parra Alonso, Ignacio; Salinas Maldonado, Carlota; Fernández Llorca, David; Sotelo Vázquez, Miguel ÁngelIdentifiers
Permanent link (URI): http://hdl.handle.net/10017/59784DOI: 10.1109/IVS.2018.8500672
ISBN: 978-1-5386-4452-2
Publisher
IEEE
Date
2018-10-22Funders
Comunidad de Madrid
European Commission
Bibliographic citation
Izquierdo Gonzalo, R., Parra Alonso, I., Salinas Maldonado, C., Fernández Llorca, D. & Sotelo Vázquez, M.A. 2018, “Semi-automatic high-accuracy labelling tool for multi-modal long-range sensor dataset”, in 2018 IEEE Intelligent Vehicles Symposium (IV), Changshu, China, pp. 1786-1791.
Keywords
Cameras
Laser radar
Calibration
Trajectory
Three-dimensional displays
Labeling
Automobiles
Description / Notes
2018 IEEE Intelligent Vehicles Symposium (IV), Changshu, Suzhou, China, June 26-30, 2018.
Project
info:eu-repo/grantAgreement/CAM/S2013-MIT-2713/ES//SEGVAUTO-TRIES-CM
info:eu-repo/grantAgreement/EC/H2020/723021/BRIDGING GAPS FOR THE ADOPTION OF AUTOMATED VEHICLES/BRAVE
Document type
info:eu-repo/semantics/bookPart
Version
info:eu-repo/semantics/acceptedVersion
Publisher's version
https://doi.org/10.1109/IVS.2018.8500672Rights
Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
© 2018 IEEE
Access rights
info:eu-repo/semantics/openAccess
Abstract
Many research works have contributed to achieve SAE levels 3 and 4 in some pre-defined areas under certain restrictions. A deeper scene understanding and precise predictions of drivers intentions are needed to continue improving autonomous driving capabilities to reach higher SAE levels. Deployment of accurate and detailed datasets could be considered as one of the most pressing needs to enhance autonomous driving capabilities. This work presents a novel data acquisition methodology for on-road vehicle trajectory collection. The proposed sensor setup improves the range and detection accuracy by using a high accuracy laser scanner covering 360° and two high-speed and high-resolution cameras. The sensor fusion increases the labelling resolution and extends the detection range sporting the best of each sensor. A Median Flow tracking algorithm and a Convolutional Neural Network enable a semi-automatic labelling process, which reduces the effort to create detailed annotated datasets. High accurate trajectories are reconstructed with few manual annotations up to 60m with a mean error below 2 cm. This methodology has been developed with a view to creating a dataset which enables the development of advanced vehicle trajectory prediction systems, and thus to contribute to human-like automated driving.
Files in this item
Files | Size | Format |
|
---|---|---|---|
Semi_automatic_Izquierdo_IEEE_ ... | 2.755Mb |
|
Files | Size | Format |
|
---|---|---|---|
Semi_automatic_Izquierdo_IEEE_ ... | 2.755Mb |
|