Current Result Document :
| 한글제목(Korean Title) |
도시 환경에서의 이미지 분할 모델 대상 적대적 물리 공격 기법 |
| 영문제목(English Title) |
Adversarial Wall: Physical Adversarial Attack on Cityscape Pretrained Segmentation Model |
| 저자(Author) |
수랸토 나우팔
라라사티 하라스타 타티마
김용수
김호원
Naufal Suryanto
Harashta Tatimma Larasati
Yongsu Kim
Howon Kim
|
| 원문수록처(Citation) |
VOL 29 NO. 02 PP. 0402 ~ 0404 (2022. 11) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Recent research has shown that deep learning models are vulnerable to adversarial attacks not only in the digital but also in the physical domain. This becomes very critical for applications that have a very high safety concern, such as self-driving cars. In this study, we propose a physical adversarial attack technique for one of the common tasks in self-driving cars, namely segmentation of the urban scene. Our method can create a texture on a wall so that it can be misclassified as a road. The demonstration of the technique on a state-of-the-art cityscape pretrained model shows a fairly high success rate, which should raise awareness of more potential attacks in self-driving cars. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
