Kinematic and linguistic interpretation of human motion via RF signal analysis
Sevgi Zübeyde Gürbüz, Evie Malaia · 2021
Chapter Contents: 14.1 Introduction 14.2 RF Sensor phenomenology and data representations 14.2.1 2D representations 14.2.2 3D representations 14.3 Deep learning for human activity recognition 14.3.1 Unsupervised pre-training versus transfer learning 14.3.2 Synthetic RF data generation with motion capture 14.3.3 Kinematic considerations in adversarial learning 14.4 Linguistic analysis of radar data for ASL recognition 14.4.1 Physical parameters of ASL signal 14.4.2 RF measurements of ASL 14.4.3 Native vs. imitation signing 14.4.4 Fractal complexity for discrimination of ASL from daily activities 14.5 ASL Recognition using RF μD signatures 14.5.1 Handcrafted features 14.5.2 Feature selection and classification results 14.6 Conclusion Acknowledgments References