Deep Learning Applications of Short-Range Radars
1st Edition
Published on 30. September 2020
358 pages
978-1-63081-747-3 (ISBN)
Description
This exciting new resource presents emerging applications of artificial intelligence and deep learning in short-range radar. The book covers applications ranging from industrial, consumer space to emerging automotive applications. The book presents several human-machine interface (HMI) applications, such as gesture recognition and sensing, human activity classification, air-writing, material classification, vital sensing, people sensing, people counting, people localization and in-cabin automotive occupancy and smart trunk opening.The underpinnings of deep learning are explored, outlining the history of neural networks and the optimization algorithms to train them. Modern deep convolutional neural network (DCNN), popular DCNN architectures for computer vision and their features are also introduced. The book presents other deep learning architectures, such as long-short term memory (LSTM), auto-encoders, variational auto-encoders (VAE), and generative adversarial networks (GAN). The application of human activity recognition as well as the application of air-writing using a network of short-range radars are outlined. This book demonstrates and highlights how deep learning is enabling several advanced industrial, consumer and in-cabin applications of short-range radars, which weren't otherwise possible. It illustrates various advanced applications, their respective challenges, and how they are been addressed using different deep learning architectures and algorithms.
More details
Other editions
Additional editions
Avik Santra | Souvik Hazra
Emerging Deep Learning Applications of Short Range Radars
Book
09/2020
Artech House Publishers
Unfortunately, price unknown
Available (delivery time upon request)
Content
- Intro
- Deep Learning Applications of Short-Range Radars
- Contents
- Preface
- 1 Introduction to Radar Signal Processing
- 1.1 Types of Radar
- 1.1.1 CW Radar
- 1.1.2 Modulated CW Radar
- 1.1.3 Impulse UWB Radar
- 1.1.4 Other Short Range Radars
- 1.2 Waveform Design and Ambiguity Function
- 1.3 System Concept
- 1.4 Target Model
- 1.5 3D Data-cube Processing
- 1.5.1 1D Processing
- 1.5.2 2D Range Doppler Images
- 1.5.3 Range Cross-Range Images
- 1.6 Detection Strategy and Clustering
- 1.6.1 Detection Algorithm
- 1.6.2 Clustering
- 1.7 Parameter Estimation and Cramer-Rao Bound
- 1.8 Tracking
- 1.8.1 Track Management
- 1.8.2 Track Filtering
- 1.9 Applications of Short-Range Radar
- 1.10 Problems
- References
- 2 Introduction to Deep Learning
- 2.1 Perceptron
- 2.2 Multilayer Perceptron
- 2.2.1 Training
- 2.2.2 Activation Functions
- 2.2.3 Optimizers
- 2.2.4 Types of Models
- 2.3 Convolutional Neural Networks
- 2.3.1 Convolution Layer
- 2.3.2 Popular Architectures
- 2.3.3 Transfer Learning
- 2.4 LSTM
- 2.5 Autoencoders
- 2.6 Variational Autoencoder
- 2.7 Generative Adversarial Network
- 2.8 Robust Deep Learning
- 2.9 Problems
- References
- 3 Gesture Sensing and Recognition
- 3.1 Introduction
- 3.1.1 RelatedWork
- 3.2 Gesture Sensing/Detection
- 3.3 Micro-Gestures
- 3.3.1 System Parameters
- 3.3.2 Micro-Gesture Set
- 3.4 2D All CNN-LSTM
- 3.4.1 Architecture and Learning
- 3.4.2 System Evaluation
- 3.5 3D CNN and Pseudo-3D CNN
- 3.5.1 3D CNN Architecture and Learning
- 3.5.2 Pseudo-3D CNN Architecture and Learning
- 3.6 Meta-Learning
- 3.6.1 Architecture and Learning
- 3.6.2 System Evaluation
- 3.7 Macro-Gestures
- 3.7.1 System Parameters
- 3.7.2 Macro-Gesture Set
- 3.8 Unguided Attention 2D DCNN-LSTM
- 3.8.1 Architecture and Learning
- 3.8.2 System Evaluation
- 3.9 FutureWork and Direction
- 3.10 Problems
- References
- 4 Human Activity Recognition and Elderly-Fall Detection
- 4.1 Introduction
- 4.1.1 RelatedWork
- 4.2 Preprocessing for Feature Image
- 4.2.1 Fast-Time FFT
- 4.2.2 Coherent Pulse Integration
- 4.2.3 MTI Filtering
- 4.2.4 Adaptive Detection Thresholding
- 4.2.5 Euclidean Clustering
- 4.2.6 Kalman Filter
- 4.3 Input Feature Images
- 4.3.1 Range Spectrogram
- 4.3.2 Doppler Spectrogram
- 4.3.3 Video of RDI
- 4.4 Human Activity Data Set
- 4.5 DCNN Activity Classification
- 4.5.1 Architecture and Learning
- 4.5.2 Results and Discussion
- 4.6 Bayesian Classification
- 4.6.1 Integrated Classifier and Tracker
- 4.6.2 Results and Discussion
- 4.7 Fall-Motion Recognition
- 4.7.1 Architecture and Learning
- 4.7.2 Deformable CNN
- 4.7.3 Loss Function
- 4.7.4 Results and Discussion
- 4.8 FutureWork and Directions
- 4.9 Problems
- References
- 5 Air-Writing
- 5.1 Introduction
- 5.2 Radar Network Placement
- 5.3 Preprocessing
- 5.3.1 Coherent Pulse Integration
- 5.3.2 Moving Target Indication Filtering
- 5.3.3 Target Detection and Selection
- 5.3.4 Localization with Trilateration
- 5.3.5 Trajectory Smoothening Filters
- 5.4 Setup and Characters
- 5.4.1 Character Set
- 5.4.2 System Parameters
- 5.4.3 Setup and Data Acquistion
- 5.5 LSTM
- 5.5.1 Architecture
- 5.5.2 Loss Function: CTC
- 5.5.3 Design Considerations
- 5.5.4 Performance Evaluation
- 5.6 Deep Convolutional Neural Networks
- 5.6.1 Architecture
- 5.6.2 Weight Initialization
- 5.6.3 Learning Schedule
- 5.6.4 Data Augmentation
- 5.6.5 Performance Evaluation
- 5.7 1D CNN-LSTM
- 5.7.1 Architecture
- 5.7.2 Performance Evaluation
- 5.8 FutureWork and Directions
- 5.9 Problems
- References
- 6 Material Classification
- 6.1 Introduction
- 6.1.1 RelatedWork
- 6.2 Features: Range Angle Images
- 6.3 Deep Convolutional Neural Networks
- 6.3.1 Architecture and Learning
- 6.3.2 Design Considerations
- 6.3.3 Results and Discussion
- 6.4 Siamese Network
- 6.4.1 Architecture and Learning
- 6.4.2 Design Considerations
- 6.4.3 Results and Discussion
- 6.5 FutureWork and Directions
- 6.6 Problems
- References
- 7 Vital Sensing and Classification
- 7.1 Introduction
- 7.2 Vital Signal Fundamentals
- 7.2.1 Preprocessing Steps
- 7.3 Heart Rate Estimation through a Deep-Learning Approach
- 7.3.1 GAN-Based Data Augmentation
- 7.3.2 Results and Discussions
- 7.4 Adaptive Signal Processing with a Tracking Approach
- 7.4.1 Algorithm
- 7.4.2 Results and Discussion
- 7.5 IQ Signal Evaluation using Deep Learning
- 7.5.1 Deep Learning Architecture
- 7.5.2 Results and Discussion
- 7.6 FutureWork and Direction
- 7.7 Problems
- References
- 8 People Sensing, Counting,and Localization
- 8.1 Introduction
- 8.2 Presence Sensing: Signal Processing Approach
- 8.2.1 Challenges
- 8.2.2 Solution
- 8.3 Presence Sensing: Deep Learning Approach
- 8.3.1 Challenges
- 8.3.2 Solution
- 8.3.3 Results and Discussion
- 8.4 People Counting: Signal Processing Approach
- 8.4.1 Challenges
- 8.4.2 Solution
- 8.5 People Counting: Deep Learning Approach
- 8.5.1 Data Preparation and Processing
- 8.5.2 Solution: Framework and Learning
- 8.5.3 Results and Discussion
- 8.6 People Detection and Localization: Signal Processing Approach
- 8.7 People Detection and Localization: Deep Learning Approach
- 8.7.1 Challenges
- 8.7.2 Architecture and Learning
- 8.7.3 Results and Discussion
- 8.8 FutureWork and Direction
- 8.9 Problems
- References
- 9 Automotive In-Cabin Sensing
- 9.1 Introduction
- 9.2 Smart Trunk Opening
- 9.2.1 Challenges
- 9.2.2 Solution
- 9.2.3 Results and Discussion
- 9.3 Vehicle Occupancy Sensing
- 9.3.1 Challenges
- 9.3.2 Solution
- 9.4 Federated Learning
- 9.4.1 Challenges
- 9.4.2 Solution
- 9.5 FutureWork and Direction
- 9.6 Problems
- References
- About the Authors
- Index
