Investigating the significance of adversarial attacks and their relation to interpretability for radar-based human activity recognition systems

Computer Science

• Interpretability 100%
• Recognition System 100%
• Human Activity Recognition 100%
• Convolutional Neural Network 100%
• Adversarial Machine Learning 100%
• Neural Network 16%
• Machine Learning 16%
• Threat Model 16%
• Cyberattack 16%
• Computer Vision 16%
• Privacy Concern 16%
• Adversarial Example 16%
• Gradient-Weighted Class Activation Mapping 16%

Keyphrases

• Human Activity Recognition Systems 100%
• Neural Network Interpretability 16%
• Gradient-based Attack 16%
• Black-box Adversarial Attack 16%
• Low-power Radar 16%