Internet of Multimedia Things (IoMT): Techniques and Applications

Contributors
Contents
1 A review on Internet of Multimedia Things (IoMT) routing protocols and quality of service
1.1 Introduction
1.2 Routing protocols in IoMT
1.2.1 Fault tolerant routing protocol
1.2.2 DDSV routing protocol
1.2.3 Optimal routing for multihop social-based D2D communications
1.2.4 Green-RPL routing protocol
1.2.5 Context-aware and load balancing RPL (CLRPL)
1.2.6 Energy-Harvesting-Aware (EHA) routing protocol
1.2.7 Optimized 3-D deployment with lifetime constraint (O3DwLC) protocol
Algorithm description
1.3 Quality of Service (QoS) routing in IoMT
1.3.1 Traffic-aware QoS routing protocol
1.3.2 QoS-aware and Heterogeneously Clustered Routing (QHCR) protocol
1.3.3 QoS in wireless multimedia sensor network based IoMT
1.3.4 SAMS framework for WMSN-based IoMT
1.4 Conclusion and future directions
References
2 Energy efficient data communication in Internet of Multimedia Things (IoMT)
2.1 Introduction
2.2 Related work
2.2.1 Routing protocol for IoMT
2.2.2 Boundary detection and virtual coordinate algorithms
2.3 System model
2.3.1 Problem description
2.4 Proposed approach
2.4.1 Working of algorithm
2.4.2 Phase I: boundary detection algorithms
2.4.3 Phase II: assignment of virtual coordinates to boundary nodes
2.4.3.1 Complexity
2.5 Implementation and results
2.5.1 Preliminaries
2.5.2 Simulation setup
2.5.2.1 Boundary and Internal nodes
2.5.2.2 Percentage of correctly detected Boundary nodes
2.5.2.3 Number of packets delivered to the sink node
2.5.2.4 Energy consumptions with respect to time
2.6 Conclusion
References
3 Visual information processing and transmission in Wireless Multimedia Sensor Networks: a deep learning based practical approach
3.1 Introduction
3.2 Literature review
3.3 Deep learning based practical approach for WMSN
3.3.1 Convolutional Neural Networks
3.3.2 Activation units
3.3.3 Max pooling
3.3.4 Batch normalization
3.3.5 Regularization
3.3.6 Transfer learning
3.3.7 Recurrent Neural Networks
3.3.8 Auto-Encoders
3.3.9 Generative Adversarial Networks
3.3.10 Mobile neural networks
3.4 Computer vision algorithms in WMSN
3.4.1 Object detection
3.4.2 Semantic segmentation
3.4.3 Image restoration, superresolution, and semantic colorization
3.5 Conclusion and future considerations
References
4 Cognitive radio-medium access control protocol for Internet of Multimedia Things (IoMT)
4.1 Introduction
4.2 Difference in Internet of Things and Internet-of-Multimedia-Things
4.2.1 Architecture
4.2.2 Performance parameters
4.2.3 Scalar data of IoT and big data of IoMT
4.3 Internet-of-Multimedia-Things and cognitive radio
4.3.1 Feasibility of cognitive radio to multimedia traffic
4.3.2 Why cognitive radio is a potential candidate for Internet-of-Multimedia-Things?
4.3.3 Cognitive radio in IoMT network
4.4 CR-MAC protocols for Internet-of-Multimedia-Things
4.5 Challenges in cognitive radio based IoMT network
4.5.1 Spectrum sensing
4.5.2 Spectrum sharing/management
4.5.3 Spectrum mobility
4.5.4 Miscellaneous challenges
4.6 Summary
References
5 Multimedia nano communication for healthcare – noise analysis
5.1 Introduction
5.2 Noise in nano communication and statistical tools
5.3 Fundamentals on various noises in MNC
5.3.1 Additive inverse Gaussian (IG) noise
5.3.2 Normal inverse Gaussian noise
5.3.3 Stable distribution noise
5.3.4 Modified Nakagami distribution noise
5.3.5 Radiation absorption noise
5.3.6 Sampling and counting noise
5.3.7 Molecular displacement noise
5.3.8 Drug delivery noise
5.3.9 Reactive obstacle noise
5.3.10 External noise
5.3.11 System noise
5.4 Fundamentals on various noises in EMNC
5.4.1 Johnson–Nyquist noise/thermal noise
5.4.2 Black-body noise
5.4.3 Doppler-shift-induced noise
5.4.4 Molecular absorption noise
5.4.5 Body radiation noise
5.5 Physical and/or stochastic models for noise in MNC
5.5.1 Additive inverse Gaussian noise (AIGN)
5.5.2 Normal inverse Gaussian noise (NIGN)
5.5.3 Stable distribution noise
5.5.3.1 Levy distribution noise
5.5.3.2 Standardized stable noise-I
5.5.3.3 Standardized stable noise-II
5.5.4 Modified Nakagami distribution noise
5.5.5 Radiation absorption noise
5.5.6 Sampling and counting noise
5.5.7 Molecular displacement noise
5.5.8 Drug delivery noise
5.5.9 Reactive obstacle noise
5.5.10 External noise
5.5.11 System noise
5.6 Physical and/or stochastic models for noise in EMNC
5.6.1 Johnson–Nyquist noise/thermal noise
5.6.2 Black-body noise
5.6.3 Doppler-shift-induced noise
5.6.4 Molecular absorption noise
5.6.5 Body radiation noise
5.7 Simulation results of different noises under nano communication
5.8 Open research challenges on noises in nano communication
5.8.1 Challenges on the noises in MNC
5.8.2 Challenges on the noises in EMNC
5.9 Summary
References
6 The use of deep learning in image analysis for the study of oncology
6.1 The difficulties in meeting demand
6.1.1 The medical imaging equipment issue
6.2 What is deep learning?
6.2.1 Neuron's and activation functions
6.2.1.1 Binary step function
6.2.1.2 Linear activation functions
6.2.1.3 Rectified linear unit
6.2.1.4 Sigmoid functions
6.2.2 Feedforward neural networks
6.2.2.1 The limitations of FFNs
6.2.3 Recurrent neural networks
6.2.3.1 Vanishing gradient problem
6.2.4 Long-short term memory model
6.2.5 Convolutional neural networks
6.3 Deep learning techniques and processes
6.3.1 Backpropagation
6.3.2 Image segmentation
6.3.3 Object localization in the study of oncology
6.4 Data difficulties
6.4.1 Stepping away from supervised-learning
6.4.2 Dimension reduction
6.4.2.1 Principal component analysis
6.4.2.2 Autoencoder
6.4.2.2.1 Stacked autoencoder
6.4.2.2.2 Sparse autoencoder
6.4.2.2.3 Denoising autoencoder
6.4.2.2.4 Variational autoencoder
6.4.2.2.5 Convolution autoencoder
6.4.3 Synthetic data
6.4.3.1 Super resolution
6.5 The current uses of deep learning image analysis
6.6 The future of deep learning image analysis in the study of oncology
6.7 Conclusion
References
7 Automatic analysis of the heart sound signal to build smart healthcare system
7.1 Introduction
7.1.1 Motivation
7.1.2 Contributions
7.1.2.1 TQWT based denoising algorithm for PCG signal
7.1.2.2 Analysis of multidomain features for PCG signal
7.1.2.3 Classification of PCG signal
7.2 Literature survey
7.2.1 Denoising algorithms
7.2.2 Segmentation algorithms
7.2.3 Feature extraction and classification algorithms
7.3 Methods and materials
7.3.1 Theoretical background about TQWT
Decomposition of the signal
Reconstruction using TQWT
7.3.2 HSMM based heart sound signal segmentation
7.3.3 Machine learning based classification algorithms
7.3.3.1 Support vector machine (SVM)
7.3.3.2 K-nearest neighbor (KNN)
7.3.3.3 Ensemble of multiple classifiers
7.4 Proposed methodology
7.4.1 Data preprocessing
7.4.2 TQWT based denoising algorithm
7.4.2.1 Decomposition of the signal using TQWT
7.4.2.2 Selection of a level with emphasized FHS
7.4.2.3 Suppression of in-band noise components
7.4.3 Segmentation using Springer's HSMM algorithm
7.4.4 Feature extraction
7.4.4.1 Time domain features
7.4.4.2 Frequency domain features
7.4.4.3 Time-frequency domain features
7.4.5 Classification of heart sound signal
7.5 Results and discussion
7.5.1 Performance evaluation metrics
7.5.2 Results using the SVM method
7.5.3 Results using KNN method
7.5.4 Results using ensemble method
7.5.5 Comparison of the proposed method with other methods
7.6 Conclusions
References
8 Efficient single image haze removal using CLAHE and Dark Channel Prior for Internet of Multimedia Things
8.1 Introduction
8.1.1 Efficient multimedia processing for IoMT
8.2 Background and related work
8.2.1 Dark Channel Prior (DCP)
8.2.2 Contrast-Limited Adaptive Histogram Equalization (CLAHE)
8.3 Analysis of adaptive contrast enhancement with DCP and its optimization
8.3.1 Adaptive contrast enhancement with DCP
8.3.2 Optimization for computational advantage
8.4 Results and discussion
8.4.1 Quality of haze removal
8.4.2 Performance assessment
8.4.3 Discussion
8.5 Conclusion
References
9 A supervised and unsupervised image quality assessment framework in real-time
9.1 Introduction
9.1.1 Motivating scenario
9.2 Related work
9.2.1 Full-reference image quality assessment
9.2.2 No-reference image quality assessment
9.3 Contributions
9.4 Definitions
9.4.1 Data model
9.4.2 Data manipulation functions
9.5 Data quality
9.5.1 Neural network architecture for image quality assessment
9.5.1.1 Supervised neural network module
9.5.1.2 Unsupervised neural network module
9.5.2 Face alignment anomaly detection
9.5.3 Image score
9.6 Framework
9.6.1 Stream processing module
9.6.2 Back-end module
9.6.2.1 Identity Recognition Module
9.6.2.1.1 Face Detection
9.6.2.1.2 Face Recognition
9.6.2.1.3 Entities Knowledge Base
9.6.2.1.4 Distortion Detection
9.6.2.1.5 Distortion Recognition
9.6.2.1.6 Distortion Trained Model
9.6.2.2 Quality Estimation Module
9.6.2.2.1 Entity Manipulation Function
9.6.2.2.2 Face Alignment Anomaly Detection
9.6.2.2.3 Neural Network Image Quality Assessment Module
9.6.2.2.4 Max Image Score
9.7 Experiments
9.7.1 Experimental setup
9.7.2 Experimental protocol
9.7.3 Results
9.7.3.1 Performances comparison
9.7.3.2 Evaluating the images quality after applying a manipulation function
9.7.3.3 Evaluating the framework in real-time
9.8 Conclusion
References
10 A computational approach to understand building floor plan images using machine learning techniques
10.1 Introduction
10.2 Motivation of the problem
10.3 Problem statement
10.3.1 Brief description of the work done
10.4 Literature survey
10.4.1 State of the art in graphic recognition
10.4.2 State of the art in floor plan analysis
10.4.3 Publicly available floor plan datasets
10.4.4 Symbol spotting in document images
10.4.5 Image description generation
10.4.6 Evaluation of text generation
10.5 Descriptive narration generation from floor plan images
10.5.1 System overview
10.5.2 Room annotation learning model
10.5.2.1 Room semantic segmentation
10.5.2.2 Decor characterization
10.5.3 Bag of decor (BoD)
10.5.3.1 Room annotation learning
10.5.4 Semistructured description generation
10.5.5 Experimental findings
10.5.5.1 Results of decor characterization
10.5.5.2 Results of room annotation learning
10.5.6 Results for description generation
10.5.6.1 Qualitative results
10.5.6.2 Quantitative analysis
10.6 Application to smart homes and buildings
10.7 Conclusion
References
Index