A SIMPLE AND UNIVERSAL ROTATION EQUIVARIANT POINT-CLOUD NETWORK

被引：0

作者：

Finkelshtein, Ben ^{[1
]}

Baskin, Chaim ^{[1
]}

Maron, Haggai ^{[2
]}

Dym, Nadav ^{[3
]}

机构：

[1] Technion, Fac Comp Sci, Haifa, Israel

[2] NVIDIA Res, Haifa, Israel

[3] Technion, Fac Math, Haifa, Israel

来源：

TOPOLOGICAL, ALGEBRAIC AND GEOMETRIC LEARNING WORKSHOPS 2022, VOL 196 | 2022年 / 196卷

关键词：

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Equivariance to permutations and rigid motions is an important inductive bias for various 3D learning problems. Recently it has been shown that the equivariant Tensor Field Network architecture is universal- it can approximate any equivariant function. In this paper we suggest a much simpler architecture, prove that it enjoys the same universality guarantees and evaluate its performance on Modelnet40. The code to reproduce our experiments is available at https://github.com/simpleinvariance/UniversalNetwork

引用

页数：9

共 50 条

[31] LANDSLIDE DETECTION BY INDICES OF LIDAR POINT-CLOUD DENSITY
Liu, Jin-King
Hsu, Wei-Chen
Yang, Mon-Shieh
Shieh, Yu-Chung
Shih, Tian-Yuan
[J]. 2010 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM, 2010, : 3960 - 3963
[32] The construction of geometric models from point-cloud data
Smith, G
Claustre, T
[J]. 16TH INTERNATIONAL CONFERENCE ON COMPUTER-AIDED PRODUCTION ENGINEERING - CAPE 2000, 2000, 2000 (05): : 3 - 11
[33] A Novel Preprocessing Method for Dynamic Point-Cloud Compression
Lee, Mun-yong
Lee, Sang-ha
Jung, Kye-dong
Lee, Seung-hyun
Kwon, Soon-chul
[J]. APPLIED SCIENCES-BASEL, 2021, 11 (13):
[34] Egomotion Estimation by Point-Cloud Back-Mapping
Geng, Haokun
Nicolescu, Radu
Klette, Reinhard
[J]. COMPUTER VISION AND GRAPHICS, ICCVG 2014, 2014, 8671 : 228 - 235
[35] Point-cloud noncontact metrology of freeform optical surfaces
Yao, Jianing
Anderson, Alexander
Rolland, Jannick P.
[J]. OPTICS EXPRESS, 2018, 26 (08): : 10242 - 10265
[36] Patch-Collaborative Spectral Point-Cloud Denoising
Rosman, G.
Dubrovina, A.
Kimmel, R.
[J]. COMPUTER GRAPHICS FORUM, 2013, 32 (08) : 1 - 12
[37] SO(3) Rotation Equivariant Point Cloud Completion using Attention-based Vector Neurons
Wu, Hang
Miao, Yubin
[J]. 2022 INTERNATIONAL CONFERENCE ON 3D VISION, 3DV, 2022, : 280 - 290
[38] Interpretable Rotation-Equivariant Quaternion Neural Networks for 3D Point Cloud Processing
Shen, Wen
Wei, Zhihua
Ren, Qihan
Zhang, Binbin
Huang, Shikun
Fan, Jiaqi
Zhang, Quanshi
[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2024, 46 (05) : 3290 - 3304
[39] Point-Cloud Data Reduction Based on Neighborhood-Point Position Feature
Zhang, Zihui
Guan, Yunlan
[J]. LASER & OPTOELECTRONICS PROGRESS, 2023, 60 (16)
[40] Robust point-cloud registration based on dense point matching and probabilistic modeling
Netto, Gustavo Marques
Oliveira, Manuel M.
[J]. VISUAL COMPUTER, 2022, 38 (9-10): : 3217 - 3230

← 1 2 3 4 5 →