Wearable Computing With Distributed Deep Learning Hierarchy: A Study of Fall Detection

Abstract

Chapter 1. Introduction 1 Chapter 2. Literature Review 5 Chapter 3. System Overview 8 Data Acquisition and Processing 9 Distributed Hierarchical Deep Learning Chapter 4. Experiment and Results Experimental Setup Result and Analysis Chapter 5. Discussion Why to use resources on the cloud server? What are the requirements about the incoming data during the training process? Why to use paired smart insole rather than only one smart insole? Chapter 6. Conclusion Chapter 7. Suggested Future Research References iv List of Tables 4.1 The main specs of the mobile device's hardware. 4.2 The comparison experimental results of using smart insoles and wearable sensors placed on SOIs for fall detection,W-waist, H-hand, T-thigh, I-smart insole, with k-nearest neighbor (knn), support vector machines (svm), decision tree (dt), discriminant (dc),Multi-Layer Perceptron (mlp) and convolutional neural network (cnn). 4.3 The best results through the six algorithm studies. The combination represents the combination of smart Insole, smartwatch and smartphone. 4.4 Performance of two different neural networks CNN and MLP under two different parameter exchange protocols. 4.5 Running time and Running Memory usage. 5.1 The confusion matrix using CNN with only one smart insole as inputs. CM-confusion matrix, Fa-Falling, SD-Sit Down, Wa-Walking, BP-Bend to Pick up, St-Standing, Be-Bending, Si-Sitting, Ly-Lying, Sq-Squatting and SP-Squat to pick up 5.2 The confusion matrix using CNN with two smart insole as inputs. v List of Figures 3.1 This figure shows the fall detection system flowchart. The sensor data collected from the smartwatch, smartphone and smart insoles. All the data are sent to the smartphone simultaneously. The smartphone will combine all three sensor data and perform deep learning through the system. During the training process, each smartphone transmits the intermediate output and label to the cloud server which will calculate corresponding gradient-outputs ∇ n j L(w, b; x, y) and release it back to the smartphone. After finishing the training process, the smartphone will deliver the accumulated gradients to the cloud server, and the cloud server will update the consensus model according to these gradients. The updated weights will distribute to one or more smartphones which send the requests. 3.2 Overview of DHNN architecture. (a) is the traditional DNN architecture. (b) introduces HNN architecture including a single smartphone and the cloud server. And (c) illustrates DHNN architecture in which the system further extends to multiple mobile deices. 4.1 (i), (ii), (iii), (iv), (v) representing the changes of acceleration, angular velocity, and orientation on smartwatch and smartphone, as well as the moving trend of pressure center of the smart insoles when presenting five different kinds of fall movements. Plot (vi), (vii), (viii), (ix), (x) showing the changes when presenting five different ADLs. vi 4.2 The distributed hierarchical deep learning architecture used in the system evaluation. The blocks in orange and blue represen...

Artificial Intelligence for the Metaverse: A Survey

“…Furthermore, a lightweight CNN with few layers in a cascade connection was designed to learn the physical activity patterns from encoded activity images. In [140], a system of fall detection using wearable devices was proposed for IoT-based healthcare services, in which a hierarchical DL framework with CNN architectures was developed for collaboratively processing sensory data at local devices and a cloud server. As capably working with multiple wearable devices (e.g., smartphone, smartwatch, and smart insoles), the system yielded high correct detection rate with high data privacy.…”

On Blockchain and IoT Integration Platforms: Current Implementation Challenges and Future Perspectives

“…Due to some of the privacy and security concerns of IoT solutions as well as the sensitivity of the data that is obtained from them, especially in areas like healthcare [1] and supply chain management [1], some current implementations have infused blockchain into their IoT systems. Some examples of these include applications in wearables [65,66], healthcare [3], data storage [67], smart transportation system, and smart cities [68].…”

Development of an Anomaly Alert System Triggered by Unusual Behaviors at Home

“…The studies that focus on this area can also be subdivided, with some of them focusing on the development of wearable systems or devices [18][19][20], while others use edge devices (most commonly mobile phones) for information processing [21][22][23]. These products are usually highly effective in detecting falls.…”