Rui Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Rui Liu has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Cuntai Guan
Nanyang Technological University
5 shared publications - Han Yu
Nanyang technological university, Singapore, Singapore
5 shared publications - Anran Li
School of Medicine, Yale University, New Haven, CT, USA
3 shared publications - Gernot R. Muller-Putz
Graz University of Technology
2 shared publications - Linlin You
School of Intelligent Systems Engineering, Shenzhen Campus of Sun Yat-sen University, Shenzhen, China
2 shared publications - Sheng Liu
Networked Systems Security Group (NSS), KTH Royal Institute of Technology, Stockholm, Sweden
2 shared publications
Latest Publications
- Open Access
- 17 January 2026
- Published by Elsevier in Neural Networks
- Vol. 198, 108608
EEG-to-gait decoding via phase-aware representation learning
- 17 January 2026
- Published by Elsevier in Neural Networks
- Vol. 198, 108608
Open AccessEEG-to-gait decoding via phase-aware representation learning
- Open Access
- 22 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 34 (15344320) , 441-454
EEG2GAIT: A Hierarchical Graph Convolutional Network for EEG-Based Gait Decoding
Decoding gait dynamics from EEG signals presents significant challenges due to the complex spatial dependencies of motor processes, the need for accurate temporal and spectral feature extraction, and the scarcity of high-quality gait EEG datasets. To address these issues, we propose EEG2GAIT, a novel hierarchical graph-based model that captures multi-level spatial embeddings of EEG channels using a Hierarchical Graph Convolutional Network (GCN) Pyramid. To further improve decoding performance, we introduce a Hybrid Temporal-Spectral Reward (HTSR) loss function, which integrates time-domain, frequency-domain, and reward-based loss components. In addition, we contribute a new Gait-EEG Dataset (GED), consisting of synchronized EEG and lower-limb joint angle data collected from 50 participants across two laboratory visits. Extensive experiments demonstrate that EEG2GAIT with HTSR achieves superior performance on the GED dataset, reaching a Pearson correlation coefficient (r) of 0.959, a coefficient of determination (R2) of 0.914, and a Mean Absolute Error (MAE) of 0.193. On the MoBI dataset, EEG2GAIT likewise consistently outperforms existing methods, achieving an r of 0.779, an R2 of 0.597, and an MAE of 4.384. Statistical analyses confirm that these improvements are significant compared to all prior models. Ablation studies further validate the contributions of the hierarchical GCN modules and the proposed HTSR loss, while saliency analysis highlights the involvement of motor-related brain regions in decoding tasks. Collectively, these findings underscore EEG2GAIT’s potential for advancing brain-computer interface applications, particularly in lower-limb rehabilitation and assistive technologies.
- 22 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 34 (15344320) , 441-454
Open AccessEEG2GAIT: A Hierarchical Graph Convolutional Network for EEG-Based Gait Decoding
Decoding gait dynamics from EEG signals presents significant challenges due to the complex spatial dependencies of motor processes, the need for accurate temporal and spectral feature extraction, and the scarcity of high-quality gait EEG datasets. To address these issues, we propose EEG2GAIT, a novel hierarchical graph-based model that captures multi-level spatial embeddings of EEG channels using a Hierarchical Graph Convolutional Network (GCN) Pyramid. To further improve decoding performance, we introduce a Hybrid Temporal-Spectral Reward (HTSR) loss function, which integrates time-domain, frequency-domain, and reward-based loss components. In addition, we contribute a new Gait-EEG Dataset (GED), consisting of synchronized EEG and lower-limb joint angle data collected from 50 participants across two laboratory visits. Extensive experiments demonstrate that EEG2GAIT with HTSR achieves superior performance on the GED dataset, reaching a Pearson correlation coefficient (r) of 0.959, a coefficient of determination (R2) of 0.914, and a Mean Absolute Error (MAE) of 0.193. On the MoBI dataset, EEG2GAIT likewise consistently outperforms existing methods, achieving an r of 0.779, an R2 of 0.597, and an MAE of 4.384. Statistical analyses confirm that these improvements are significant compared to all prior models. Ablation studies further validate the contributions of the hierarchical GCN modules and the proposed HTSR loss, while saliency analysis highlights the involvement of motor-related brain regions in decoding tasks. Collectively, these findings underscore EEG2GAIT’s potential for advancing brain-computer interface applications, particularly in lower-limb rehabilitation and assistive technologies.
- 10 December 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (2) , 1-17
FedIBD: a federated learning framework in asynchronous mode for imbalanced data
- 10 December 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (2) , 1-17
FedIBD: a federated learning framework in asynchronous mode for imbalanced data
- Top Cited PapersOpen Access
- 21 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (3) , 1909-1918
EEG-Deformer: A Dense Convolutional Transformer for Brain-Computer Interfaces
Effectively learning the temporal dynamics in electroencephalogram (EEG) signals is challenging yet essential for decoding brain activities using brain-computer interfaces (BCIs). Although Transformers are popular for their long-term sequential learning ability in the BCI field, most methods combining Transformers with convolutional neural networks (CNNs) fail to capture the coarse-to-fine temporal dynamics of EEG signals. To overcome this limitation, we introduce EEG-Deformer, which incorporates two main novel components into a CNN-Transformer: (1) a Hierarchical Coarse-to-Fine Transformer (HCT) block that integrates a Fine-grained Temporal Learning (FTL) branch into Transformers, effectively discerning coarse-to-fine temporal patterns; and (2) a Dense Information Purification (DIP) module, which utilizes multi-level, purified temporal information to enhance decoding accuracy. Comprehensive experiments on three representative cognitive tasks-cognitive attention, driving fatigue, and mental workload detection-consistently confirm the generalizability of our proposed EEG-Deformer, demonstrating that it either outperforms or performs comparably to existing state-of-the-art methods. Visualization results show that EEG-Deformer learns from neurophysiologically meaningful brain regions for the corresponding cognitive tasks. The source code can be found at https://github.com/yi-ding-cs/EEG-Deformer .
- 21 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (3) , 1909-1918
Top Cited PapersOpen AccessEEG-Deformer: A Dense Convolutional Transformer for Brain-Computer Interfaces
Effectively learning the temporal dynamics in electroencephalogram (EEG) signals is challenging yet essential for decoding brain activities using brain-computer interfaces (BCIs). Although Transformers are popular for their long-term sequential learning ability in the BCI field, most methods combining Transformers with convolutional neural networks (CNNs) fail to capture the coarse-to-fine temporal dynamics of EEG signals. To overcome this limitation, we introduce EEG-Deformer, which incorporates two main novel components into a CNN-Transformer: (1) a Hierarchical Coarse-to-Fine Transformer (HCT) block that integrates a Fine-grained Temporal Learning (FTL) branch into Transformers, effectively discerning coarse-to-fine temporal patterns; and (2) a Dense Information Purification (DIP) module, which utilizes multi-level, purified temporal information to enhance decoding accuracy. Comprehensive experiments on three representative cognitive tasks-cognitive attention, driving fatigue, and mental workload detection-consistently confirm the generalizability of our proposed EEG-Deformer, demonstrating that it either outperforms or performs comparably to existing state-of-the-art methods. Visualization results show that EEG-Deformer learns from neurophysiologically meaningful brain regions for the corresponding cognitive tasks. The source code can be found at https://github.com/yi-ding-cs/EEG-Deformer .
- Top Cited PapersOpen Access
- 8 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 36 (3) , 4279-4295
Federated Graph Neural Networks: Overview, Techniques, and Challenges
Graph neural networks (GNNs) have attracted extensive research attention in recent years due to their capability to progress with graph data and have been widely used in practical applications. As societies become increasingly concerned with the need for data privacy protection, GNNs face the need to adapt to this new normal. Besides, as clients in federated learning (FL) may have relationships, more powerful tools are required to utilize such implicit information to boost performance. This has led to the rapid development of the emerging research field of federated GNNs (FedGNNs). This promising interdisciplinary field is highly challenging for interested researchers to grasp. The lack of an insightful survey on this topic further exacerbates the entry difficulty. In this article, we bridge this gap by offering a comprehensive survey of this emerging field. We propose a 2-D taxonomy of the FedGNN literature: 1) the main taxonomy provides a clear perspective on the integration of GNNs and FL by analyzing how GNNs enhance FL training as well as how FL assists GNN training and 2) the auxiliary taxonomy provides a view on how FedGNNs deal with heterogeneity across FL clients. Through discussions of key ideas, challenges, and limitations of existing works, we envision future research directions that can help build more robust, explainable, efficient, fair, inductive, and comprehensive FedGNNs.
- 8 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 36 (3) , 4279-4295
Top Cited PapersOpen AccessFederated Graph Neural Networks: Overview, Techniques, and Challenges
Graph neural networks (GNNs) have attracted extensive research attention in recent years due to their capability to progress with graph data and have been widely used in practical applications. As societies become increasingly concerned with the need for data privacy protection, GNNs face the need to adapt to this new normal. Besides, as clients in federated learning (FL) may have relationships, more powerful tools are required to utilize such implicit information to boost performance. This has led to the rapid development of the emerging research field of federated GNNs (FedGNNs). This promising interdisciplinary field is highly challenging for interested researchers to grasp. The lack of an insightful survey on this topic further exacerbates the entry difficulty. In this article, we bridge this gap by offering a comprehensive survey of this emerging field. We propose a 2-D taxonomy of the FedGNN literature: 1) the main taxonomy provides a clear perspective on the integration of GNNs and FL by analyzing how GNNs enhance FL training as well as how FL assists GNN training and 2) the auxiliary taxonomy provides a view on how FedGNNs deal with heterogeneity across FL clients. Through discussions of key ideas, challenges, and limitations of existing works, we envision future research directions that can help build more robust, explainable, efficient, fair, inductive, and comprehensive FedGNNs.
- 7 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (8) , 9659-9674
AFM3D: An Asynchronous Federated Meta-Learning Framework for Driver Distraction Detection
Driver Distraction Detection (3D) is of great significance in helping intelligent vehicles decide whether to remind drivers or take over the driving task and avoid traffic accidents. However, the current centralized learning paradigm of 3D has become unpractical because of rising limitations on data sharing and increasing concerns about user privacy. In this context, 3D is further facing three emerging challenges, namely data islands, data heterogeneity, and the straggler issue. To jointly address these three issues and make the 3D model training and deployment more practical and efficient, this paper proposes an Asynchronous Federated Meta-learning framework called AFM3D. Specifically, AFM3D bridges data islands through Federated Learning (FL), a novel distributed learning paradigm that enables multiple clients (i.e., private vehicles with individual data of drivers) to learn a global model collaboratively without data exchange. Moreover, AFM3D further utilizes meta-learning to tackle data heterogeneity by training a meta-model that can adapt to new driver data quickly with satisfactory performance. Finally, AFM3D is designed to operate in an asynchronous mode to reduce delays caused by stragglers and achieve efficient learning. A temporally weighted aggregation strategy is also designed to handle stale models commonly encountered in the asynchronous mode and in turn, optimize the aggregation direction. Extensive experiment results show that AFM3D can boost performance in terms of model accuracy, recall, F1 score, test loss, and learning speed by 7.61%, 7.44%, 7.95%, 9.95%, and 50.91%, respectively, against five state-of-the-art methods.
- 7 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (8) , 9659-9674
AFM3D: An Asynchronous Federated Meta-Learning Framework for Driver Distraction Detection
Driver Distraction Detection (3D) is of great significance in helping intelligent vehicles decide whether to remind drivers or take over the driving task and avoid traffic accidents. However, the current centralized learning paradigm of 3D has become unpractical because of rising limitations on data sharing and increasing concerns about user privacy. In this context, 3D is further facing three emerging challenges, namely data islands, data heterogeneity, and the straggler issue. To jointly address these three issues and make the 3D model training and deployment more practical and efficient, this paper proposes an Asynchronous Federated Meta-learning framework called AFM3D. Specifically, AFM3D bridges data islands through Federated Learning (FL), a novel distributed learning paradigm that enables multiple clients (i.e., private vehicles with individual data of drivers) to learn a global model collaboratively without data exchange. Moreover, AFM3D further utilizes meta-learning to tackle data heterogeneity by training a meta-model that can adapt to new driver data quickly with satisfactory performance. Finally, AFM3D is designed to operate in an asynchronous mode to reduce delays caused by stragglers and achieve efficient learning. A temporally weighted aggregation strategy is also designed to handle stale models commonly encountered in the asynchronous mode and in turn, optimize the aggregation direction. Extensive experiment results show that AFM3D can boost performance in terms of model accuracy, recall, F1 score, test loss, and learning speed by 7.61%, 7.44%, 7.95%, 9.95%, and 50.91%, respectively, against five state-of-the-art methods.
- Open Access
- 9 January 2024
- Published by Elsevier in Neural Networks
- Vol. 172, 106100
Aggregating intrinsic information to enhance BCI performance through federated learning
- 9 January 2024
- Published by Elsevier in Neural Networks
- Vol. 172, 106100
Open AccessAggregating intrinsic information to enhance BCI performance through federated learning
- 27 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 53, 1005-1009
Extreme Learning Machine Classifier Based on Novel Particle Swarm Optimization Algorithm
Aiming at the problem of low classification accuracy caused by random input weights and hidden layer bias in ELM (Extreme Learning Machines, ELM), a method based on IPSO (Improved Particle Swarm Optimization, IPSO) to optimize the network classification model of Extreme Learning Machines is proposed. Firstly, the nonlinear inertia weight and dynamic acceleration factor are used to balance the exploration and mining of the PSO algorithm; Secondly, the variables of the ELM model are enhanced by IPSO to reduce the influence caused by strong volatility; Finally, simulation results show that compared with SVM, ELM, GA-ELM, IPSO-ELM could effectively improve the precision of classification prediction.
- 27 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 53, 1005-1009
Extreme Learning Machine Classifier Based on Novel Particle Swarm Optimization Algorithm
Aiming at the problem of low classification accuracy caused by random input weights and hidden layer bias in ELM (Extreme Learning Machines, ELM), a method based on IPSO (Improved Particle Swarm Optimization, IPSO) to optimize the network classification model of Extreme Learning Machines is proposed. Firstly, the nonlinear inertia weight and dynamic acceleration factor are used to balance the exploration and mining of the PSO algorithm; Secondly, the variables of the ELM model are enhanced by IPSO to reduce the influence caused by strong volatility; Finally, simulation results show that compared with SVM, ELM, GA-ELM, IPSO-ELM could effectively improve the precision of classification prediction.
- 29 March 2023
- Published by SPIE-Intl Soc Optical Eng
- Vol. 12566, 396-402
FSC-based data fusion sensitivity assessment scheme
In recent years, with the large amount of data generated, data fusion and sharing have become more and more important. However, the expansion of the scope of data fusion and sharing has led to the increasing problem of data leakage risk. In this paper, we improve the FSC fusion algorithm to evaluate the sensitivity change during data fusion from three aspects: data set, data items and data values, so that the fused data sensitivity is more accurate, and later the fused data is anonymized for sharing. The algorithm reduces information loss and provides better privacy protection.
- 29 March 2023
- Published by SPIE-Intl Soc Optical Eng
- Vol. 12566, 396-402
FSC-based data fusion sensitivity assessment scheme
In recent years, with the large amount of data generated, data fusion and sharing have become more and more important. However, the expansion of the scope of data fusion and sharing has led to the increasing problem of data leakage risk. In this paper, we improve the FSC fusion algorithm to evaluate the sensitivity change during data fusion from three aspects: data set, data items and data values, so that the fused data sensitivity is more accurate, and later the fused data is anonymized for sharing. The algorithm reduces information loss and provides better privacy protection.
- Open Access
Towards Interpretable Federated Learning
Federated learning (FL) enables multiple data owners to build machine learning models collaboratively without exposing their private local data. In order for FL to achieve widespread adoption, it is important to balance the need for performance, privacy-preservation and interpretability, especially in mission critical applications such as finance and healthcare. Thus, interpretable federated learning (IFL) has become an emerging topic of research attracting significant interest from the academia and the industry alike. Its interdisciplinary nature can be challenging for new researchers to pick up. In this paper, we bridge this gap by providing (to the best of our knowledge) the first survey on IFL. We propose a unique IFL taxonomy which covers relevant works enabling FL models to explain the prediction results, support model debugging, and provide insights into the contributions made by individual data owners or data samples, which in turn, is crucial for allocating rewards fairly to motivate active and reliable participation in FL. We conduct comprehensive analysis of the representative IFL approaches, the commonly adopted performance evaluation metrics, and promising directions towards building versatile IFL techniques.
Open AccessTowards Interpretable Federated Learning
Federated learning (FL) enables multiple data owners to build machine learning models collaboratively without exposing their private local data. In order for FL to achieve widespread adoption, it is important to balance the need for performance, privacy-preservation and interpretability, especially in mission critical applications such as finance and healthcare. Thus, interpretable federated learning (IFL) has become an emerging topic of research attracting significant interest from the academia and the industry alike. Its interdisciplinary nature can be challenging for new researchers to pick up. In this paper, we bridge this gap by providing (to the best of our knowledge) the first survey on IFL. We propose a unique IFL taxonomy which covers relevant works enabling FL models to explain the prediction results, support model debugging, and provide insights into the contributions made by individual data owners or data samples, which in turn, is crucial for allocating rewards fairly to motivate active and reliable participation in FL. We conduct comprehensive analysis of the representative IFL approaches, the commonly adopted performance evaluation metrics, and promising directions towards building versatile IFL techniques.