Point Cloud Augmentation with Weighted Local Transformations

Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation is a standard approach to compensate for the scarcity of data, it has been less explored in the point cloud literature. In this paper, we prop...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings / IEEE International Conference on Computer Vision S. 528 - 537
Hauptverfasser: Kim, Sihyeon, Lee, Sanghyeok, Hwang, Dasol, Lee, Jaewon, Hwang, Seong Jae, Kim, Hyunwoo J.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.10.2021
Schlagworte:
ISSN:2380-7504
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation is a standard approach to compensate for the scarcity of data, it has been less explored in the point cloud literature. In this paper, we propose a simple and effective augmentation method called PointWOLF for point cloud augmentation. The proposed method produces smoothly varying non-rigid deformations by locally weighted transformations centered at multiple anchor points. The smooth deformations allow diverse and realistic augmentations. Furthermore, in order to minimize the manual efforts to search the optimal hyperparameters for augmentation, we present AugTune, which generates augmented samples of desired difficulties producing targeted confidence scores. Our experiments show our framework consistently improves the performance for both shape classification and part segmentation tasks. Particularly, with PointNet++, PointWOLF achieves the state-of-the-art 89.7 accuracy on shape classification with the real-world ScanObjectNN dataset. The code is available at https://github.com/mlvlab/PointWOLF.
AbstractList Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation is a standard approach to compensate for the scarcity of data, it has been less explored in the point cloud literature. In this paper, we propose a simple and effective augmentation method called PointWOLF for point cloud augmentation. The proposed method produces smoothly varying non-rigid deformations by locally weighted transformations centered at multiple anchor points. The smooth deformations allow diverse and realistic augmentations. Furthermore, in order to minimize the manual efforts to search the optimal hyperparameters for augmentation, we present AugTune, which generates augmented samples of desired difficulties producing targeted confidence scores. Our experiments show our framework consistently improves the performance for both shape classification and part segmentation tasks. Particularly, with PointNet++, PointWOLF achieves the state-of-the-art 89.7 accuracy on shape classification with the real-world ScanObjectNN dataset. The code is available at https://github.com/mlvlab/PointWOLF.
Author Kim, Sihyeon
Lee, Sanghyeok
Kim, Hyunwoo J.
Lee, Jaewon
Hwang, Seong Jae
Hwang, Dasol
Author_xml – sequence: 1
  givenname: Sihyeon
  surname: Kim
  fullname: Kim, Sihyeon
  email: sh_bs15@korea.ac.kr
  organization: Korea University
– sequence: 2
  givenname: Sanghyeok
  surname: Lee
  fullname: Lee, Sanghyeok
  email: cat0626@korea.ac.kr
  organization: Korea University
– sequence: 3
  givenname: Dasol
  surname: Hwang
  fullname: Hwang, Dasol
  email: dd_sol@korea.ac.kr
  organization: Korea University
– sequence: 4
  givenname: Jaewon
  surname: Lee
  fullname: Lee, Jaewon
  email: 2j1ejyu@korea.ac.kr
  organization: Korea University
– sequence: 5
  givenname: Seong Jae
  surname: Hwang
  fullname: Hwang, Seong Jae
  email: sjh95@pitt.edu
  organization: University of Pittsburgh
– sequence: 6
  givenname: Hyunwoo J.
  surname: Kim
  fullname: Kim, Hyunwoo J.
  email: hyunwoojkim@korea.ac.kr
  organization: Korea University
BookMark eNotzE9LwzAYgPEoCm5zn0AP-QKdb_42AS-jOB0U9DD1ON426RZpE2kzxG8vqKfn8uOZk4uYoifklsGKMbB326p6k8ZyvuLA2QoAlD0jS1saprWS3DCuzsmMCwNFqUBekfk0fQAIy42ekfuXFGKmVZ9Ojq5Ph8HHjDmkSL9CPtJ3Hw7H7B2tU4s93Y0Ypy6Nwy-Zrsllh_3kl_9dkNfNw656Kurnx221rovAQeQCVaOFdMw7aJnDxqFuO-G88w1KEMqZRnVl07JOOjRWIyrtuBROYcuh9WJBbv6-wXu__xzDgOP33pYMJAPxA5zfTHM
CODEN IEEPAD
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/ICCV48922.2021.00059
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISBN 9781665428125
1665428120
EISSN 2380-7504
EndPage 537
ExternalDocumentID 9710410
Genre orig-research
GroupedDBID 29O
6IE
6IF
6IH
6IK
6IL
6IM
6IN
AAJGR
AAWTH
ACGFS
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IPLJI
M43
OCL
RIE
RIL
RIO
RNS
ID FETCH-LOGICAL-i203t-a5b634d1ed0c1dabda6cf3dedeba4035d8b5f7bc1f4da896aa56d243d5ac20ce3
IEDL.DBID RIE
ISICitedReferencesCount 42
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000797698900053&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:25:37 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-a5b634d1ed0c1dabda6cf3dedeba4035d8b5f7bc1f4da896aa56d243d5ac20ce3
PageCount 10
ParticipantIDs ieee_primary_9710410
PublicationCentury 2000
PublicationDate 2021-Oct.
PublicationDateYYYYMMDD 2021-10-01
PublicationDate_xml – month: 10
  year: 2021
  text: 2021-Oct.
PublicationDecade 2020
PublicationTitle Proceedings / IEEE International Conference on Computer Vision
PublicationTitleAbbrev ICCV
PublicationYear 2021
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0039286
Score 2.4842467
Snippet Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation...
SourceID ieee
SourceType Publisher
StartPage 528
SubjectTerms grouping and shape
Manuals
Neural networks
Point cloud compression
Recognition and classification
Segmentation
Shape
Task analysis
Three-dimensional displays
Training
Vision applications and systems
Title Point Cloud Augmentation with Weighted Local Transformations
URI https://ieeexplore.ieee.org/document/9710410
WOSCitedRecordID wos000797698900053&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFA5zePA0dRN_k4NHu6Xta5qAFymKgowdpu428uNVBtrK1vr3m3R1KnjxFgIh8CUv38tLvvcIuQCuNDriCJQECECYMBAmsYG7PEPO4pyFaJtiE-l4LGYzOemQy40WBhGbz2c49M3mLd-WpvahspF0dAheT7WVpnyt1fo6dR3NC95K40ImR_dZ9gRCRl5rFYXDJg_JrwIqDX_c9v438y4ZfAvx6GRDMXukg8U-6bWeI23tctUnV5NyUVQ0ey1rS6_rl7dWUVRQH2elz03804148MxFpz-cVbfpBuTx9maa3QVtXYRgEbG4ClSieQzWochMaJW2ips8tmhRK2BxYoVO8lSbMAerhORKJdxGENvE52M0GB-QblEWeEgoMrCQguRonDULLpRA7sAUOddCKTgifQ_G_H2d-mLe4nD8d_cJ2fFor_-6nZJutazxjGybj2qxWp436_UJwbOYHA
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NSwMxEB1KFfRUtRW_zcGj2-7HbJoFL7JYWqylh6q9lWwyKwXdSrvr7zfZrlXBi7cQCIGXTN5kkjcDcIVcJmSIw5ERooNCeY5QoXbM5RlTN0hdj3RZbKI7GonpNBrX4HqjhSGi8vMZtW2zfMvXC1XYUFknMnSIVk-1ZStnVWqtr3PXEL3glTjOc6POII6fUES-VVv5XrvMRPKrhErJIL3G_-beg9a3FI-NNySzDzXKDqBR-Y6sssxVE27Gi3mWs_h1UWh2W7y8VZqijNlIK3suI6BmxNByF5v8cFfNtmvBY-9uEvedqjKCM_fdIHdkmPAAtcHRVZ6WiZZcpYEmTYlENwi1SMK0mygvRS1FxKUMufYx0KHNyKgoOIR6tsjoCBi5qLGLESdl7FlwIQVxA6ZIeSKkxGNoWjBm7-vkF7MKh5O_uy9hpz95GM6Gg9H9Kexa5Nc_386gni8LOodt9ZHPV8uLcu0-Ad36m2U
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+%2F+IEEE+International+Conference+on+Computer+Vision&rft.atitle=Point+Cloud+Augmentation+with+Weighted+Local+Transformations&rft.au=Kim%2C+Sihyeon&rft.au=Lee%2C+Sanghyeok&rft.au=Hwang%2C+Dasol&rft.au=Lee%2C+Jaewon&rft.date=2021-10-01&rft.pub=IEEE&rft.eissn=2380-7504&rft.spage=528&rft.epage=537&rft_id=info:doi/10.1109%2FICCV48922.2021.00059&rft.externalDocID=9710410