A5107207899- Adversarial Robustness in Machine Learning
- Advanced SAR Imaging Techniques
- Higher Education and Teaching Methods
- Multimodal Machine Learning Applications
- Chromium effects and bioremediation
- ZnO doping and properties
- High-Velocity Impact and Material Behavior
- Neural Networks and Applications
- Acoustic Wave Resonator Technologies
- Gas Sensing Nanomaterials and Sensors
- Smart Grid and Power Systems
- Magnetic Properties and Applications
- Education and Work Dynamics
- Computational Drug Discovery Methods
- Topic Modeling
- Advanced Decision-Making Techniques
- Diverse Interdisciplinary Research Innovations
- Evaluation Methods in Various Fields
- Wireless Signal Modulation Classification
- Radiation Detection and Scintillator Technologies
- Advanced Image Processing Techniques
- Metal Extraction and Bioleaching
- Bacillus and Francisella bacterial research
- Advanced Image and Video Retrieval Techniques
- Geoscience and Mining Technology
PLA Information Engineering University
2024
Lanzhou University of Technology
2018
Anhui Normal University
2010
Sichuan University
2010
Institute of Engineering
2004
Synthetic aperture radar automatic target recognition (SAR-ATR) models based on deep neural networks (DNNs) are vulnerable to attacks of adversarial examples. Universal attack algorithms can help evaluate and improve the robustness SAR-ATR have become a research hotspot. However, current universal limitations. First, considering difficulty in obtaining information attacking models, there is an urgent need design algorithm under black-box scenario. Second, given acquiring SAR images,...
With the rapid development of artificial intelligence technology, deep learning has achieved significant advantages in synthetic aperture radar automatic target recognition (SAR-ATR). However, previous research showed that addition small perturbations not easily detected by human eye can lead to SAR-ATR model errors; is, they are affected adversarial attacks. To solve problem long computation time existing SAR sparse attack algorithms, we propose a fast (FSAA) algorithm. First, an end-to-end...
Abstract Existing synthetic aperture radar (SAR) adversarial attack algorithms primarily focus on the digital image domain, and constructing examples in real‐world scenarios presents significant challenging hurdles. This study proposes template‐based universal (TUAA) algorithm. Initially, a SAR interference template generator module is constructed to derive perturbation domain. The designed loss function guides parameter updating of generator, thereby improving effectiveness concealment....
Crystallized solid-solution films, Sr1-x Ca x MoO4 and Ba (0 ≤ 1), have been prepared on Si substrates by chemical solution processing. The structure composition of the films were investigated using X-ray diffraction Vegard's law. No obvious phase separation was observed when annealed at 650°C. peaks position dependent content Ca, Sr, in respectively. And different growing habits caused element species lead to oriented growth films.