GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation

Min Seok Lee; Seok Woo Yang; Sung Won Han

doi:10.1109/WACV56688.2023.00065

GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation

Min Seok Lee^*
, Seok Woo Yang
, Sung Won Han^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

30 Citations (Scopus)

Abstract

While point cloud semantic segmentation is a significant task in 3D scene understanding, this task demands a time-consuming process of fully annotating labels. To address this problem, recent studies adopt a weakly supervised learning approach under the sparse annotation. Different from the existing studies, this study aims to reduce the epistemic uncertainty measured by the entropy for a precise semantic segmentation. We propose the graphical information gain based attention network called GaIA, which alleviates the entropy of each point based on the reliable information. The graphical information gain discriminates the reliable point by employing relative entropy between target point and its neighborhoods. We further introduce anchor-based additive angular margin loss, ArcPoint. The ArcPoint optimizes the unlabeled points containing high entropy towards semantically similar classes of the labeled points on hypersphere space. Experimental results on S3DIS and ScanNet-v2 datasets demonstrate our framework outperforms the existing weakly supervised methods.

Original language	English
Title of host publication	Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	582-591
Number of pages	10
ISBN (Electronic)	9781665493468
DOIs	https://doi.org/10.1109/WACV56688.2023.00065
Publication status	Published - 2023
Event	23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023 - Waikoloa, United States Duration: 2023 Jan 3 → 2023 Jan 7

Publication series

Name	Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023

Conference

Conference	23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023
Country/Territory	United States
City	Waikoloa
Period	23/1/3 → 23/1/7

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Keywords

Algorithms: 3D computer vision
Image recognition and understanding (object detection, categorization, segmentation, scene modeling, visual reasoning)

ASJC Scopus subject areas

Artificial Intelligence
Computer Science Applications
Computer Vision and Pattern Recognition

Access to Document

10.1109/WACV56688.2023.00065

Cite this

Lee, M. S., Woo Yang, S., & Han, S. W. (2023). GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation. In Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023 (pp. 582-591). (Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WACV56688.2023.00065

GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation. / Lee, Min Seok; Woo Yang, Seok; Han, Sung Won.
Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 582-591 (Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Lee, MS, Woo Yang, S & Han, SW 2023, GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation. in Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023. Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023, Institute of Electrical and Electronics Engineers Inc., pp. 582-591, 23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023, Waikoloa, United States, 23/1/3. https://doi.org/10.1109/WACV56688.2023.00065

Lee MS, Woo Yang S, Han SW. GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation. In Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 582-591. (Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023). doi: 10.1109/WACV56688.2023.00065

Lee, Min Seok ; Woo Yang, Seok ; Han, Sung Won. / GaIA : Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation. Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 582-591 (Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023).

@inproceedings{f4b4c83ccfab49d99c05fae6f9f07e47,

title = "GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation",

abstract = "While point cloud semantic segmentation is a significant task in 3D scene understanding, this task demands a time-consuming process of fully annotating labels. To address this problem, recent studies adopt a weakly supervised learning approach under the sparse annotation. Different from the existing studies, this study aims to reduce the epistemic uncertainty measured by the entropy for a precise semantic segmentation. We propose the graphical information gain based attention network called GaIA, which alleviates the entropy of each point based on the reliable information. The graphical information gain discriminates the reliable point by employing relative entropy between target point and its neighborhoods. We further introduce anchor-based additive angular margin loss, ArcPoint. The ArcPoint optimizes the unlabeled points containing high entropy towards semantically similar classes of the labeled points on hypersphere space. Experimental results on S3DIS and ScanNet-v2 datasets demonstrate our framework outperforms the existing weakly supervised methods.",

keywords = "Algorithms: 3D computer vision, Image recognition and understanding (object detection, categorization, segmentation, scene modeling, visual reasoning)",

author = "Lee, \{Min Seok\} and \{Woo Yang\}, Seok and Han, \{Sung Won\}",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023 ; Conference date: 03-01-2023 Through 07-01-2023",

year = "2023",

doi = "10.1109/WACV56688.2023.00065",

language = "English",

series = "Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "582--591",

booktitle = "Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023",

}

TY - GEN

T1 - GaIA

T2 - 23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023

AU - Lee, Min Seok

AU - Woo Yang, Seok

AU - Han, Sung Won

PY - 2023

Y1 - 2023

N2 - While point cloud semantic segmentation is a significant task in 3D scene understanding, this task demands a time-consuming process of fully annotating labels. To address this problem, recent studies adopt a weakly supervised learning approach under the sparse annotation. Different from the existing studies, this study aims to reduce the epistemic uncertainty measured by the entropy for a precise semantic segmentation. We propose the graphical information gain based attention network called GaIA, which alleviates the entropy of each point based on the reliable information. The graphical information gain discriminates the reliable point by employing relative entropy between target point and its neighborhoods. We further introduce anchor-based additive angular margin loss, ArcPoint. The ArcPoint optimizes the unlabeled points containing high entropy towards semantically similar classes of the labeled points on hypersphere space. Experimental results on S3DIS and ScanNet-v2 datasets demonstrate our framework outperforms the existing weakly supervised methods.

AB - While point cloud semantic segmentation is a significant task in 3D scene understanding, this task demands a time-consuming process of fully annotating labels. To address this problem, recent studies adopt a weakly supervised learning approach under the sparse annotation. Different from the existing studies, this study aims to reduce the epistemic uncertainty measured by the entropy for a precise semantic segmentation. We propose the graphical information gain based attention network called GaIA, which alleviates the entropy of each point based on the reliable information. The graphical information gain discriminates the reliable point by employing relative entropy between target point and its neighborhoods. We further introduce anchor-based additive angular margin loss, ArcPoint. The ArcPoint optimizes the unlabeled points containing high entropy towards semantically similar classes of the labeled points on hypersphere space. Experimental results on S3DIS and ScanNet-v2 datasets demonstrate our framework outperforms the existing weakly supervised methods.

KW - Algorithms: 3D computer vision

KW - Image recognition and understanding (object detection, categorization, segmentation, scene modeling, visual reasoning)

UR - https://www.scopus.com/pages/publications/85149003151

U2 - 10.1109/WACV56688.2023.00065

DO - 10.1109/WACV56688.2023.00065

M3 - Conference contribution

AN - SCOPUS:85149003151

T3 - Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023

SP - 582

EP - 591

BT - Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 3 January 2023 through 7 January 2023

ER -

GaIA: Graphical Information Gain based Attention Network for Weakly Supervised Point Cloud Semantic Segmentation

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this