메뉴 건너뛰기
.. 내서재 .. 알림
소속 기관/학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
로그인 회원가입 고객센터 ENG
주제분류

추천
검색
질문

논문 기본 정보

자료유형
학술저널
저자정보
Hong Joo Lee (Technical University of Munich) Kyo Seok Lee (Chung-Ang University) Hak Gu Kim (Chung-Ang University)
저널정보
중앙대학교 영상콘텐츠융합연구소 TECHART: Journal of Arts and Imaging Science TECHART: Journal of Arts and Imaging Science Vol.12 No.1
발행연도
2025.1
수록면
27 - 34 (8page)
DOI
10.15323/techart.2025.1.12.1.27

이용수

표지
📌
연구주제
📖
연구배경
🔬
연구방법
🏆
연구결과
AI에게 요청하기
추천
검색
질문

초록· 키워드

오류제보하기
Previous adversarial attacks on three-dimensional (3D) point clouds were perturbed by considering the shape to maintain the structures of the original point cloud. However, these often resulted in detectable perturbed points, making them vulnerable to adversarial defense attacks. To address this, in this study, we introduce a novel adversarial point cloud perturbation approach by employing cage-based 3D deformation to naturally deform the geometry of 3D point clouds while preserving the adversarial attack power. Unlike previous methods that aimed to preserve the original shape, the proposed method strategically transforms the structure of the point cloud, thereby enabling effective and robust attacks against advanced defense mechanisms. Cage-based 3D deformation was utilized to control the cage surrounding the point cloud, allowing coordinated transformations by propagating changes from the cage vertices to internal points. This ensures cohesive and smooth deformations that are ideal for generating effective adversarial examples. Extensive experiments on ModelNet40 and ShapeNet datasets validated its effectiveness, demonstrating its ability to evade traditional defenses while providing natural- looking shapes. This study advances the understanding of the relationship between 3D deformation and adversarial point-cloud perturbations.

목차

Abstract
1. Introduction
2. Related works
3. Proposed method
4. Experiments
5. Conclusion
References

참고문헌 (0)

참고문헌 신청

함께 읽어보면 좋을 논문

논문 유사도에 따라 DBpia 가 추천하는 논문입니다. 함께 보면 좋을 연관 논문을 확인해보세요!

최근 본 자료

전체보기

댓글(0)

0

UCI(KEPA) : I410-151-25-02-092443438