ASFM-Net: Asymmetrical Siamese Feature Matching Network for Point Completion

Yaqi Xia, Yan Xia, Wei Li, Rui Song, Kailang Cao, Uwe Stilla

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

32 Zitate (Scopus)

Abstract

We tackle the problem of object completion from point clouds and propose a novel point cloud completion network employing an Asymmetrical Siamese Feature Matching strategy, termed as ASFM-Net. Specifically, the Siamese auto-encoder neural network is adopted to map the partial and complete input point cloud into a shared latent space, which can capture detailed shape prior. Then we design an iterative refinement unit to generate complete shapes with fine-grained details by integrating prior information. Experiments are conducted on the PCN dataset and the Completion3D benchmark, demonstrating the state-of-the-art performance of the proposed ASFM-Net. Our method achieves the 1st place in the leaderboard of Completion3D and outperforms existing methods with a large margin, about 12%. The codes and trained models are released publicly at https://github.com/Yan-Xia/ASFM-Net.

OriginalspracheEnglisch
TitelMM 2021 - Proceedings of the 29th ACM International Conference on Multimedia
Herausgeber (Verlag)Association for Computing Machinery, Inc
Seiten1938-1947
Seitenumfang10
ISBN (elektronisch)9781450386517
DOIs
PublikationsstatusVeröffentlicht - 17 Okt. 2021
Veranstaltung29th ACM International Conference on Multimedia, MM 2021 - Virtual, Online, China
Dauer: 20 Okt. 202124 Okt. 2021

Publikationsreihe

NameMM 2021 - Proceedings of the 29th ACM International Conference on Multimedia

Konferenz

Konferenz29th ACM International Conference on Multimedia, MM 2021
Land/GebietChina
OrtVirtual, Online
Zeitraum20/10/2124/10/21

Fingerprint

Untersuchen Sie die Forschungsthemen von „ASFM-Net: Asymmetrical Siamese Feature Matching Network for Point Completion“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren