Reducing domain shift in synthetic data augmentation for semantic segmentation of 3D point clouds - Equipe Robot interaction, Ambient system, Machine learning, Behaviour, Optimization Access content directly
Conference Papers Year : 2022

Reducing domain shift in synthetic data augmentation for semantic segmentation of 3D point clouds

Abstract

The use of deep learning in semantic segmentation of point clouds enables a drastic improvement of segmentation precision. However, available datasets are restrained to a few applications with limited applicability to other fields. Using synthetic and real data can alleviate the burden of creating a dedicated dataset at the cost of domain-shift that is mostly addressed during training, while treating the problem directly on the data has been less explored. Towards this goal, two methods to alleviate domain shift are proposed, firstly by enhanced generation and sampling of synthetic data and secondly by leveraging color information of unlabeled point clouds to color synthetic, uncoloured data. Obtained results confirm their usefulness in improving semantic segmentation result (+3.43 into mIoU for a network trained on S3DIS zone 1). More importantly, the devised coloring method shows the ability of a point-based network to link color information with recurrent geometric features. Finally, the presented methods are able to bridge the domain-shift gap even in cases where inclusion of raw synthetic data during training impedes learning.
Fichier principal
Vignette du fichier
SMC22_0391_FI.pdf (5.32 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03796618 , version 1 (04-10-2022)

Identifiers

Cite

Romain Cazorla, Line Poinel, Panagiotis Papadakis, Cédric Buche. Reducing domain shift in synthetic data augmentation for semantic segmentation of 3D point clouds. IEEE International Conference on Systems, Man, and Cybernetics (SMC), Oct 2022, Prague, Czech Republic. ⟨10.1109/SMC53654.2022.9945480⟩. ⟨hal-03796618⟩
47 View
95 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More