Deep Transfer Learning with Graph Neural Network for Sensor-Based Human Activity Recognition

Yan, Yan; Liao, Tianzheng; Zhao, Jinjin; Wang, Jiahong; Li, Ma; Lv, Wei; Xiong, Jie; Wang, Lei

doi:10.48550/arxiv.2203.07910

Cited by 4 publications

(3 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GNN is a neural network that operates on a graph structure to produce desired results. In wearable WBANs, they are being used for human action recognition [ 237 , 238 ].…”

Section: Open Issuesmentioning

confidence: 99%

A Survey on Wireless Wearable Body Area Networks: A Perspective of Technology and Economy

Bhatti

Saleem

Imran

et al. 2022

Sensors

View full text Add to dashboard Cite

The deployment of wearable or body-worn devices is increasing rapidly, and thus researchers’ interests mainly include technical and economical issues, such as networking, interoperability, security, power optimization, business growth and regulation. To address these issues properly, previous survey papers usually focused on describing the wireless body area network architecture and network protocols. This implies that deployment issues and awareness issues of wearable and BAN devices are not emphasized in previous work. To defeat this problem, in this study, we have focused on feasibility, limitations, and security concerns in wireless body area networks. In the aspect of the economy, we have focused on the compound annual growth rate of these devices in the global market, different regulations of wearable/wireless body area network devices in different regions and countries of the world and feasible research projects for wireless body area networks. In addition, this study focuses on the domain of devices that are equally important to physicians, sportsmen, trainers and coaches, computer scientists, engineers, and investors. The outcomes of this study relating to physicians, fitness trainers and coaches indicate that the use of these devices means they would be able to treat their clients in a more effective way. The study also converges the focus of businessmen on the Annual Growth Rate (CAGR) and provides manufacturers and vendors with information about different regulatory bodies that are monitoring and regulating WBAN devices. Therefore, by providing deployment issues in the aspects of technology and economy at the same time, we believe that this survey can serve as a preliminary material that will lead to more advancements and improvements in deployment in the area of wearable wireless body area networks. Finally, we present open issues and further research direction in the area of wireless body area networks.

show abstract

“…GNN is a neural network that operates on a graph structure to produce desired results. In wearable WBANs, they are being used for human action recognition [ 237 , 238 ].…”

Section: Open Issuesmentioning

confidence: 99%

A Survey on Wireless Wearable Body Area Networks: A Perspective of Technology and Economy

Bhatti

Saleem

Imran

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Video-based HARs, in particular, have been widely examined over the past decade, with excellent outcomes in every case [29]. This is because each video in the video-based technique is made up of numerous frames, each of which contains information on the subject's movement while also keeping track of time.…”

Section: Har Has Been An Active Area Of Research In Computer Vision A...mentioning

confidence: 99%

Ensembled Transfer Learning Based Multichannel Attention Networks for Human Activity Recognition in Still Images

et al. 2022

View full text Add to dashboard Cite

Human activity recognition is one of the most difficult tasks in computer vision. Due to the lack of time information, detecting human activities from still photos is more difficult than sensor-based or videobased techniques. Recently, various deep learning based solutions are being proposed one after another, and their performance is constantly improving. In this paper, we proposed a convolutional neural architecture by ensembling transfer learning based multi-channel attention networks. Here, four CNN branches were used to make feature fusion based ensembling and in each branch, an attention module was used to extract the contextual information from the feature map produced by existing pre-trained models. Finally, the extracted feature maps from four branches were concatenated and fed to fully connected network to produce the final recognition output. We considered 3 different datasets, Stanford 40 actions, BU-101 and Willow human actions datasets to evaluate our system. Experimental analysis showed that the proposed ensembled convolutional architecture outperformed previous works by a noteworthy margin.

show abstract

“…When working with heterogeneous data, relative inductive bias can be introduced through guiding models in learning dependencies between sensors. Several works in HAR use wearable sensor data and skeleton data that exemplify this: variants of spatial–temporal graph convolution network [ 44 , 45 ] and variants of residual graph convolutional networks [ 46 , 47 ]. While graph convolution networks have been successful in HAR for wearables, we show in this work that they tend not to be as successful as attention-based graph neural networks in HAR for smart homes, especially since attention-based models are able to leverage the prior knowledge, indicating that some neighbors might be more informative than others.…”

Section: Introductionmentioning

confidence: 99%

Using Graphs to Perform Effective Sensor-Based Human Activity Recognition in Smart Homes

Plötz

2024

Sensors

View full text Add to dashboard Cite

There has been a resurgence of applications focused on human activity recognition (HAR) in smart homes, especially in the field of ambient intelligence and assisted-living technologies. However, such applications present numerous significant challenges to any automated analysis system operating in the real world, such as variability, sparsity, and noise in sensor measurements. Although state-of-the-art HAR systems have made considerable strides in addressing some of these challenges, they suffer from a practical limitation: they require successful pre-segmentation of continuous sensor data streams prior to automated recognition, i.e., they assume that an oracle is present during deployment, and that it is capable of identifying time windows of interest across discrete sensor events. To overcome this limitation, we propose a novel graph-guided neural network approach that performs activity recognition by learning explicit co-firing relationships between sensors. We accomplish this by learning a more expressive graph structure representing the sensor network in a smart home in a data-driven manner. Our approach maps discrete input sensor measurements to a feature space through the application of attention mechanisms and hierarchical pooling of node embeddings. We demonstrate the effectiveness of our proposed approach by conducting several experiments on CASAS datasets, showing that the resulting graph-guided neural network outperforms the state-of-the-art method for HAR in smart homes across multiple datasets and by large margins. These results are promising because they push HAR for smart homes closer to real-world applications.

show abstract

Deep Transfer Learning with Graph Neural Network for Sensor-Based Human Activity Recognition

Cited by 4 publications

References 50 publications

A Survey on Wireless Wearable Body Area Networks: A Perspective of Technology and Economy

A Survey on Wireless Wearable Body Area Networks: A Perspective of Technology and Economy

Ensembled Transfer Learning Based Multichannel Attention Networks for Human Activity Recognition in Still Images

Using Graphs to Perform Effective Sensor-Based Human Activity Recognition in Smart Homes

Contact Info

Product

Resources

About