17 - 22 July, 2022 • Physical: Kuala Lumpur Convention Centre (KLCC), Kuala Lumpur, Malaysia • Hybrid: Online Platform

Preserving Our Heritage, Enabling Our Future Through Remote Sensing

17 - 22 July, 2022 • Physical: Kuala Lumpur Convention Centre (KLCC), Kuala Lumpur, Malaysia • Hybrid: Online Platform

TU5.V6.1

A UNIVERSAL ADVERSARIAL ATTACK ON CNN-SAR IMAGE CLASSIFICATION BY FEATURE DICTIONARY MODELING

Weibo Qin, Feng Wang, Fudan University, China
Session:
SAR Image Classification Oral Virtual
Track:
Data Analysis Methods (Optical, Multispectral,Hyperspectral, SAR)
Location:
Virtual Room 6
Presentation Time:
Tue, 19 Jul, 19:50 - 20:05 Malaysia Time (UTC +8)
Tue, 19 Jul, 13:50 - 14:05 Central European Time (UTC +1)
Tue, 19 Jul, 11:50 - 12:05 UTC
Tue, 19 Jul, 07:50 - 08:05 Eastern Time (UTC -5)
Session Co-Chairs:
Andrea Buono, Università degli Studi di Napoli Parthenope and David A. Clausi, University of Waterloo
Session Manager:
Moon Khiang Chin
Presentation
Discussion
Resources
No resources available.
Session TU5.V6
TU5.V6.1: A UNIVERSAL ADVERSARIAL ATTACK ON CNN-SAR IMAGE CLASSIFICATION BY FEATURE DICTIONARY MODELING
Weibo Qin, Feng Wang, Fudan University, China
TU5.V6.2: SEMI-SUPERVISED SEA ICE CLASSIFICATION OF SAR IMAGERY BASED ON GRAPH CONVOLUTIONAL NETWORK
Mingzhe Jiang, Xinwei Chen, Linlin Xu, David A. Clausi, University of Waterloo, Canada
TU5.V6.3: AN IMPROVED DECOMPOSITION ALGORITHM TO DIFFERENTIATE FOREST FROM URBAN TARGETS WITH STRONG DOUBLE SCATTERING IN POLSAR DATA
Yin Zhang, Dingfeng Duan, University of Electronic Science and Technology of China, China; Yong Wang, Hong Li, East Carolina University, United States
TU5.V6.4: DEPTHWISE SEPARABLE RESIDUAL NETWORK BASED ON UNET FOR POLSAR IMAGES CLASSIFICATION
Wen Xie, Ruonan Wang, Xin Yang, Wenqiang Hua, Xi’an University of Posts and Telecommunications, China
TU5.V6.5: A KNOWLEDGE DISTILLATION METHOD BASED ON IQE ATTENTION MECHANISM FOR TARGET RECOGNITION IN SAR IMAGERY
Jielei Wang, Zongyong Cui, Zongjie Cao, Changjie Cao, University of Electronic Science and Technology of China, China; Ting Jiang, Megvii Technology Limited, China
TU5.V6.6: SEMANTIC SEGMENTATION OF SAR IMAGES THROUGH FULLY CONVOLUTIONAL NETWORKS AND HIERARCHICAL PROBABILISTIC GRAPHICAL MODELS
Martina Pastorino, University of Genova, INRIA d'Université Côte d'Azur, Italy; Gabriele Moser, Sebastiano Serpico, University of Genova, Italy; Josiane Zerubia, INRIA d'Université Côte d'Azur, France