S-NET: End-to-end Unsupervised Learning of Long-Term 3D Stable objects

Hroob, Ibrahim, Molina Mellado, Sergio, Polvara, Riccardo , Cielniak, Grzegorz and Hanheide, Marc (2023) S-NET: End-to-end Unsupervised Learning of Long-Term 3D Stable objects. In: 18th International Conference on Intelligent Autonomous Systems, 4 July—7 July, Suwon, Korea.

S-NET: End-to-end Unsupervised Learning of Long-Term 3D Stable objects
Author's Accepted Manuscript

Request a copy
[img] PDF
IAS2023_ICRA2023_COLR2022 (2).pdf - Whole Document
Restricted to Repository staff only

Item Type:Conference or Workshop contribution (Paper)
Item Status:Live Archive


In this research, we present an end-to-end data-driven pipeline for determining the long-term stability status of objects within a given environment, specifically distinguishing between static and dynamic objects. Understanding object stability is key for mobile robots since longterm stable objects can be exploited as landmarks for long-term localisation. Our pipeline includes a labelling method that utilizes historical data from the environment to generate training data for a neural network. Rather than utilizing discrete labels, we propose the use of point-wise continuous label values, indicating the spatio-temporal stability of individual points, to train a point cloud regression network named S-NET. Our approach is evaluated on point cloud data from two parking lots in the NCLT dataset, and the results show that our proposed solution, outperforms direct training of a classification model for static vs dynamic object classification.

Keywords:Mobile robotics, SLAM, point cloud
Subjects:G Mathematical and Computer Sciences > G760 Machine Learning
H Engineering > H671 Robotics
G Mathematical and Computer Sciences > G740 Computer Vision
Divisions:College of Science > School of Computer Science
ID Code:54690
Deposited On:23 Jun 2023 11:16

Repository Staff Only: item control page