Naslov
Challenging the Universal Representation of Deep Models for 3D Point Cloud Registration

Autori
Bojanic, David ; Bartol, Kristijan ; Forest, Josep ; Gumhold, Stefan ; Petkovic, Tomislav ; Tomislav, Pribanic

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the British Machine Vision Conference (BMVA) 2022 Workshop: Universal Representations for Computer Vision / - London : Delhi : BMVA Press, 2022, 1-15

Skup
33rd British Machine Vision Conference 2022

Mjesto i datum
London, Ujedinjeno Kraljevstvo, 21.11.2022. - 24.11.2022

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
deep model, 3D registration, iterative closest point, rotation matrix, 3DMatch. KITTI, ETH

Sažetak
Learning universal representations across different applications domain is an ungoing research problem. In fact, finding universal architecture within the same application but across different types of datasets is still unsolved problem too, especially in applications involving processing 3D point clouds. In this work we experimentally test several stateof-the-art learning-based methods for 3D point cloud registration against the proposed non-learning baseline registration method. The proposed method either outperforms or achieves comparable results w.r.t. learning based methods. In addition, we propose a dataset on which learning based methods have a hard time to generalize. Our proposed method and dataset, along with the provided experiments, can be used in further research in studying effective solutions for universal representations. Our source code is available at: github.com/DavidBoja/greedy-grid- search

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo

POVEZANOST RADA