A Shortcut Enhanced LSTM-GCN Network for Multi-Sensor Based Human Motion Tracking

Li, Xiaoyu; Ye, Chaoxiang; Huang, Binhua; Zhou, Zhenning; Su, Yuanzhe; Ma, Yue; Yi, Zhengkun; Wu, Xinyu

doi:10.1109/tase.2023.3307890

Cited by 4 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even if these methods support the analysis of multivariate data generated at multiple nodes, an important limitation is that they extract temporal correlations from low-level features in adjacent nodes through an initial modeling step. By doing so, the subsequent operations, which act on top of the extracted high-level features, do not take into account spatial correlations in low-level layers, which are partially unexploited or entirely lost, resulting in a performance degradation on the subsequent downstream task [39], [40], [42], [43]. Some effort has been devoted to mitigate this issue with the introduction of skip connections that propagate spatial information [39], [43] as well as spatial memory cells [40].…”

Section: B Neural Network For Temporal and Spatial Datamentioning

confidence: 99%

“…By doing so, the subsequent operations, which act on top of the extracted high-level features, do not take into account spatial correlations in low-level layers, which are partially unexploited or entirely lost, resulting in a performance degradation on the subsequent downstream task [39], [40], [42], [43]. Some effort has been devoted to mitigate this issue with the introduction of skip connections that propagate spatial information [39], [43] as well as spatial memory cells [40]. Other limitations include the lack of exploitation of graph-based relationships among nodes, and the lack of model interpretability.…”

Section: B Neural Network For Temporal and Spatial Datamentioning

confidence: 99%

“…However, one common issue is that spatial patterns are only extracted on low-level features, i.e., the input data representation, before LSTM-based modeling. This aspect may result in a limited ability for deeper layers to exploit more complex spatial correlations, therefore putting more emphasis on temporal information [39], [40], [41], [42], [43].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

GAP-LSTM: Graph-Based Autocorrelation Preserving Networks for Geo-Distributed Forecasting

Altieri,

Corizzo,

Ceci

2024

IEEE Trans. Neural Netw. Learning Syst.

View full text Add to dashboard Cite

Forecasting methods are important decision support tools in geo-distributed sensor networks. However, challenges such as the multivariate nature of data, the existence of multiple nodes, and the presence of spatio-temporal autocorrelation increase the complexity of the task. Existing forecasting methods are unable to address these challenges in a combined manner, resulting in a suboptimal model accuracy. In this article, we propose GAP-LSTM, a novel geo-distributed forecasting method that leverages the synergic interaction of graph convolution, attention-based long short-term memory (LSTM), 2-D-convolution, and latent memory states to effectively exploit spatio-temporal autocorrelation in multivariate data generated by multiple nodes, resulting in improved modeling capabilities. Our extensive evaluation, involving real-world datasets on traffic, energy, and pollution domains, showcases the ability of our method to outperform state-of-the-art forecasting methods. An ablation study confirms that all method components provide a positive contribution to the accuracy of the extracted forecasts. The method also provides an interpretable visualization that complements forecasts with additional insights for domain experts.

show abstract

Section: B Neural Network For Temporal and Spatial Datamentioning

confidence: 99%

Section: B Neural Network For Temporal and Spatial Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GAP-LSTM: Graph-Based Autocorrelation Preserving Networks for Geo-Distributed Forecasting

Altieri,

Corizzo,

Ceci

2024

IEEE Trans. Neural Netw. Learning Syst.

View full text Add to dashboard Cite

show abstract

“…In terms of visual sensors [ 5 ], human body movements can be predicted by modeling the spatial information of the body’s skeletal points. Common models used for predicting human actions include convolutional neural network (CNN) models [ 6 ] and prediction models based on graph convolutional networks (GCN) [ 7 ]. Study [ 8 ] uses a multi-path convolutional network to learn the movement trajectory features of each joint in the human body.…”

Section: Introduction and Related Workmentioning

confidence: 99%

A Novel Prediction Method of Transfer-Assisted Action Oriented to Individual Differences for the Excretion Care Robot

Wang,

Hao,

et al. 2023

Sensors

View full text Add to dashboard Cite

The excretion care robot’s (ECR) accurate recognition of transfer-assisted actions is crucial during its usage. However, transfer action recognition is a challenging task, especially since the differentiation of actions seriously affects its recognition speed, robustness, and generalization ability. We propose a novel approach for transfer action recognition assisted by a bidirectional long- and short-term memory (Bi-LSTM) network combined with a multi-head attention mechanism. Firstly, we utilize posture sensors to detect human movements and establish a lightweight three-dimensional (3D) model of the lower limbs. In particular, we adopt a discrete extended Kalman filter (DEKF) to improve the accuracy and foresight of pose solving. Then, we construct an action prediction model that incorporates a fused Bi-LSTM with Multi-head attention (MHA Bi-LSTM). The MHA extracts key information related to differentiated movements from different dimensions and assigns varying weights. Utilizing the Bi-LSTM network effectively combines past and future information to enhance the prediction results of differentiated actions. Finally, comparisons were made by three subjects in the proposed method and with two other time series based neural network models. The reliability of the MHA Bi-LSTM method was verified. These experimental results show that the introduced MHA Bi-LSTM model has a higher accuracy in predicting posture sensor-based excretory care actions. Our method provides a promising approach for handling transfer-assisted action individual differentiation in excretion care tasks.

show abstract

Research on cascading fault-guided search of high-voltage lines based on GCN-LSTM

Du,

2024

2024 6th Asia Energy and Electrical Engineering Symposium (AEEES)

View full text Add to dashboard Cite

A Shortcut Enhanced LSTM-GCN Network for Multi-Sensor Based Human Motion Tracking

Cited by 4 publications

References 32 publications

GAP-LSTM: Graph-Based Autocorrelation Preserving Networks for Geo-Distributed Forecasting

GAP-LSTM: Graph-Based Autocorrelation Preserving Networks for Geo-Distributed Forecasting

A Novel Prediction Method of Transfer-Assisted Action Oriented to Individual Differences for the Excretion Care Robot

Research on cascading fault-guided search of high-voltage lines based on GCN-LSTM

Contact Info

Product

Resources

About