Xuesong Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xuesong Li has a total of 89 co-authors. Here are the top co-authors by number of shared publications:
- Yifan Zhu
Beijing University of Posts and Telecommunications, Beijing, China
8 shared publications - Gen Shi
Beihang University
6 shared publications Ruihong Shang
School of Computer Science and Technology, Beijing Institute of Technology
4 shared publicationsWeibin Liao
Beijing Inst Technol, Sch Comp Sci & Technol, Beijing 100081, Peoples R China
3 shared publications- Weibin Liao
Peking University, Beijing, China
3 shared publications - Wenjin Liu
Department of Nephrology, Northern Jiangsu People's Hospital, Yangzhou University, Yangzhou, China
3 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (12) , 14228-14235
DGSG-Mind: Dynamic 3D Gaussian Scene Graphs for Long-Term Scene Understanding and Grounding
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (12) , 14228-14235
DGSG-Mind: Dynamic 3D Gaussian Scene Graphs for Long-Term Scene Understanding and Grounding
- 13 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (6) , 6815-6822
Dynamic Legged Ball Manipulation on Rugged Terrains With Hierarchical Reinforcement Learning
Achieving reliable object manipulation while traversing complex terrains is the missing link between agile quadruped locomotion and practical autonomy. Specifically, using traditional end-to-end reinforcement learning (RL) for dynamic ball manipulation in rugged environments presents two key challenges. The first is coordinating distinct motion modalities to integrate terrain traversal and ball control seamlessly. The second is overcoming sparse rewards in end-to-end RL, which impedes efficient policy convergence. To address these challenges, we propose a hierarchical RL framework. A high-level policy, informed by proprioceptive data and ball position, adaptively switches between pre-trained low-level skills such as ball dribbling and rough terrain navigation. We further propose Dynamic Skill-Focused Policy Optimization to suppress gradients from inactive skills and enhance critical skill learning. Both simulation and real-world experiments validate that our method outperforms baseline approaches in dynamic ball manipulation across rugged terrains, highlighting its effectiveness in challenging environments.
- 13 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (6) , 6815-6822
Dynamic Legged Ball Manipulation on Rugged Terrains With Hierarchical Reinforcement Learning
Achieving reliable object manipulation while traversing complex terrains is the missing link between agile quadruped locomotion and practical autonomy. Specifically, using traditional end-to-end reinforcement learning (RL) for dynamic ball manipulation in rugged environments presents two key challenges. The first is coordinating distinct motion modalities to integrate terrain traversal and ball control seamlessly. The second is overcoming sparse rewards in end-to-end RL, which impedes efficient policy convergence. To address these challenges, we propose a hierarchical RL framework. A high-level policy, informed by proprioceptive data and ball position, adaptively switches between pre-trained low-level skills such as ball dribbling and rough terrain navigation. We further propose Dynamic Skill-Focused Policy Optimization to suppress gradients from inactive skills and enhance critical skill learning. Both simulation and real-world experiments validate that our method outperforms baseline approaches in dynamic ball manipulation across rugged terrains, highlighting its effectiveness in challenging environments.
- 13 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 37 (8) , 4664-4676
Heterogeneous Graph-Based Multimodal Brain Network Learning
Graph neural networks (GNNs) provide powerful insights into brain neuroimaging technology from the view of graphical networks. However, most existing GNN-based models treat the brain connectome, derived from neuroimaging, as a homogeneous graph characterized by uniform node and edge types. In fact, emerging studies have reported and emphasized the significance of heterogeneity among human brain activities, especially between the two cerebral hemispheres. Thus, homogeneous-structured brain network-based graph methods are insufficient for modeling complicated cerebral activity states. To overcome this problem, we introduce a novel heterogeneous graph neural network (HeBrainGNN) for multimodal brain neuroimaging fusion learning. HeBrainGNN first conceptualizes the brain network as a heterogeneous graph with multiple types of nodes (representing the left and right hemispheres) and edges (categorizing intra- and interhemispheric interactions). We further develop a self-supervised pretraining strategy for this heterogeneous network to address the potential overfitting problem caused by the conflict between a large parameter size and a small medical data sample size. Empirical results show the superiority of the proposed model over other existing methods in brain-related disease prediction tasks. Ablation experiments show that our heterogeneous graph-based model attaches more importance to hemispheric connections that may be neglected due to their low strength by previous homogeneous graph models. Additional experiments reveal that our pretraining strategy not only addresses the challenge of limited labeled data but also significantly enhances accuracy, affirming the potential of our approach in advancing neuroimaging analysis.
- 13 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 37 (8) , 4664-4676
Heterogeneous Graph-Based Multimodal Brain Network Learning
Graph neural networks (GNNs) provide powerful insights into brain neuroimaging technology from the view of graphical networks. However, most existing GNN-based models treat the brain connectome, derived from neuroimaging, as a homogeneous graph characterized by uniform node and edge types. In fact, emerging studies have reported and emphasized the significance of heterogeneity among human brain activities, especially between the two cerebral hemispheres. Thus, homogeneous-structured brain network-based graph methods are insufficient for modeling complicated cerebral activity states. To overcome this problem, we introduce a novel heterogeneous graph neural network (HeBrainGNN) for multimodal brain neuroimaging fusion learning. HeBrainGNN first conceptualizes the brain network as a heterogeneous graph with multiple types of nodes (representing the left and right hemispheres) and edges (categorizing intra- and interhemispheric interactions). We further develop a self-supervised pretraining strategy for this heterogeneous network to address the potential overfitting problem caused by the conflict between a large parameter size and a small medical data sample size. Empirical results show the superiority of the proposed model over other existing methods in brain-related disease prediction tasks. Ablation experiments show that our heterogeneous graph-based model attaches more importance to hemispheric connections that may be neglected due to their low strength by previous homogeneous graph models. Additional experiments reveal that our pretraining strategy not only addresses the challenge of limited labeled data but also significantly enhances accuracy, affirming the potential of our approach in advancing neuroimaging analysis.
- Open Access
DynamicGSG: Dynamic 3D Gaussian Scene Graphs for Environment Adaptation
In real-world scenarios, environment changes caused by human or agent activities make it extremely challenging for robots to perform various long-term tasks. Recent works typically struggle to effectively understand and adapt to dynamic environments due to the inability to update their environment representations in memory according to environment changes and lack of fine-grained reconstruction of the environments. To address these challenges, we propose DynamicGSG, a dynamic, high-fidelity, open-vocabulary scene graph construction system leveraging Gaussian splatting. DynamicGSG builds hierarchical scene graphs using advanced vision language models to represent the spatial and semantic relationships between objects in the environments, utilizes a joint feature loss we designed to supervise Gaussian instance grouping while optimizing the Gaussian maps, and locally updates the Gaussian scene graphs according to real environment changes for long-term environment adaptation. Experiments and ablation studies demonstrate the performance and efficacy of our proposed method in terms of semantic segmentation, language-guided object retrieval, and reconstruction quality. Furthermore, we validate the dynamic updating capabilities of our system in real laboratory environments. The source code and supplementary experimental materials will be released at:~\href{https://github.com/GeLuzhou/Dynamic-GSG}{https://github.com/GeLuzhou/Dynamic-GSG}.
Open AccessDynamicGSG: Dynamic 3D Gaussian Scene Graphs for Environment Adaptation
In real-world scenarios, environment changes caused by human or agent activities make it extremely challenging for robots to perform various long-term tasks. Recent works typically struggle to effectively understand and adapt to dynamic environments due to the inability to update their environment representations in memory according to environment changes and lack of fine-grained reconstruction of the environments. To address these challenges, we propose DynamicGSG, a dynamic, high-fidelity, open-vocabulary scene graph construction system leveraging Gaussian splatting. DynamicGSG builds hierarchical scene graphs using advanced vision language models to represent the spatial and semantic relationships between objects in the environments, utilizes a joint feature loss we designed to supervise Gaussian instance grouping while optimizing the Gaussian maps, and locally updates the Gaussian scene graphs according to real environment changes for long-term environment adaptation. Experiments and ablation studies demonstrate the performance and efficacy of our proposed method in terms of semantic segmentation, language-guided object retrieval, and reconstruction quality. Furthermore, we validate the dynamic updating capabilities of our system in real laboratory environments. The source code and supplementary experimental materials will be released at:~\href{https://github.com/GeLuzhou/Dynamic-GSG}{https://github.com/GeLuzhou/Dynamic-GSG}.
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Information Systems
- Vol. 43 (1) , 1-26
RevGNN: Negative Sampling Enhanced Contrastive Graph Learning for Academic Reviewer Recommendation
Acquiring reviewers for academic submissions is a challenging recommendation scenario. Recent graph learning-driven models have made remarkable progress in the field of recommendation, but their performance in the academic reviewer recommendation task may suffer from a significant false negative issue. This arises from the assumption that unobserved edges represent negative samples. In fact, the mechanism of anonymous review results in inadequate exposure of interactions between reviewers and submissions, leading to a higher number of unobserved interactions compared to those caused by reviewers declining to participate. Therefore, investigating how to better comprehend the negative labeling of unobserved interactions in academic reviewer recommendations is a significant challenge. This study aims to tackle the ambiguous nature of unobserved interactions in academic reviewer recommendations. Specifically, we propose an unsupervised Pseudo Neg-Label strategy to enhance graph contrastive learning (GCL) for recommending reviewers for academic submissions, which we call RevGNN. RevGNN utilizes a two-stage encoder structure that encodes both scientific knowledge and behavior using Pseudo Neg-Label to approximate review preference. Extensive experiments on three real-world datasets demonstrate that RevGNN outperforms all baselines across four metrics. Additionally, detailed further analyses confirm the effectiveness of each component in RevGNN.
- 4 November 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Information Systems
- Vol. 43 (1) , 1-26
Open AccessRevGNN: Negative Sampling Enhanced Contrastive Graph Learning for Academic Reviewer Recommendation
Acquiring reviewers for academic submissions is a challenging recommendation scenario. Recent graph learning-driven models have made remarkable progress in the field of recommendation, but their performance in the academic reviewer recommendation task may suffer from a significant false negative issue. This arises from the assumption that unobserved edges represent negative samples. In fact, the mechanism of anonymous review results in inadequate exposure of interactions between reviewers and submissions, leading to a higher number of unobserved interactions compared to those caused by reviewers declining to participate. Therefore, investigating how to better comprehend the negative labeling of unobserved interactions in academic reviewer recommendations is a significant challenge. This study aims to tackle the ambiguous nature of unobserved interactions in academic reviewer recommendations. Specifically, we propose an unsupervised Pseudo Neg-Label strategy to enhance graph contrastive learning (GCL) for recommending reviewers for academic submissions, which we call RevGNN. RevGNN utilizes a two-stage encoder structure that encodes both scientific knowledge and behavior using Pseudo Neg-Label to approximate review preference. Extensive experiments on three real-world datasets demonstrate that RevGNN outperforms all baselines across four metrics. Additionally, detailed further analyses confirm the effectiveness of each component in RevGNN.
- Open Access
RevGNN: Negative Sampling Enhanced Contrastive Graph Learning for Academic Reviewer Recommendation
Acquiring reviewers for academic submissions is a challenging recommendation scenario. Recent graph learning-driven models have made remarkable progress in the field of recommendation, but their performance in the academic reviewer recommendation task may suffer from a significant false negative issue. This arises from the assumption that unobserved edges represent negative samples. In fact, the mechanism of anonymous review results in inadequate exposure of interactions between reviewers and submissions, leading to a higher number of unobserved interactions compared to those caused by reviewers declining to participate. Therefore, investigating how to better comprehend the negative labeling of unobserved interactions in academic reviewer recommendations is a significant challenge. This study aims to tackle the ambiguous nature of unobserved interactions in academic reviewer recommendations. Specifically, we propose an unsupervised Pseudo Neg-Label strategy to enhance graph contrastive learning (GCL) for recommending reviewers for academic submissions, which we call RevGNN. RevGNN utilizes a two-stage encoder structure that encodes both scientific knowledge and behavior using Pseudo Neg-Label to approximate review preference. Extensive experiments on three real-world datasets demonstrate that RevGNN outperforms all baselines across four metrics. Additionally, detailed further analyses confirm the effectiveness of each component in RevGNN.
Open AccessRevGNN: Negative Sampling Enhanced Contrastive Graph Learning for Academic Reviewer Recommendation
Acquiring reviewers for academic submissions is a challenging recommendation scenario. Recent graph learning-driven models have made remarkable progress in the field of recommendation, but their performance in the academic reviewer recommendation task may suffer from a significant false negative issue. This arises from the assumption that unobserved edges represent negative samples. In fact, the mechanism of anonymous review results in inadequate exposure of interactions between reviewers and submissions, leading to a higher number of unobserved interactions compared to those caused by reviewers declining to participate. Therefore, investigating how to better comprehend the negative labeling of unobserved interactions in academic reviewer recommendations is a significant challenge. This study aims to tackle the ambiguous nature of unobserved interactions in academic reviewer recommendations. Specifically, we propose an unsupervised Pseudo Neg-Label strategy to enhance graph contrastive learning (GCL) for recommending reviewers for academic submissions, which we call RevGNN. RevGNN utilizes a two-stage encoder structure that encodes both scientific knowledge and behavior using Pseudo Neg-Label to approximate review preference. Extensive experiments on three real-world datasets demonstrate that RevGNN outperforms all baselines across four metrics. Additionally, detailed further analyses confirm the effectiveness of each component in RevGNN.
- 6 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 12 (5) , 3627-3636
Contrastive Hierarchical Augmentation Learning for Modeling Cognitive and Multimodal Brain Network
Brain networks generated by functional magnetic resonance imaging (fMRI) have shown promising performance in characterizing cerebral social cognition and disorders. However, the scarcity of labeled data has hindered the application of deep graph learning in brain network analysis, accelerating the usage of extra label-free self-supervised contrastive graph learning. However, existing augmentation strategies commonly used in contrastive learning (CL), such as edge and node drop, do not fully benefit brain network learning due to the distribution differences between different modalities of brain neural imaging. To this end, we introduce a novel approach namely spatial–temporal hierarchical augmentation-based contrastive learning (ST-HACL) to enhance the representation learning of functional brain networks. ST-HACL leverages augmentation methods tailored specifically to brain networks. Our method employs an augmentation strategy from both spatial and temporal level during the brain network construction process to generate contrastive samples, enabling label-free self-supervised learning. We evaluate the performance of our approach on the orthostatic hypotension dataset (OH) and the Alzheimer’s disease neuroimaging initiative dataset (ADNI). Results demonstrate that our model surpasses existing graph neural network (GNN) models and graph-CL methods, achieving F1 scores of 80.61% in OH and 73.01% in ADNI. To the best of our knowledge, our study represents the first attempt at applying brain network-specific contrastive augmentation learning to fMRI analysis.
- 6 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 12 (5) , 3627-3636
Contrastive Hierarchical Augmentation Learning for Modeling Cognitive and Multimodal Brain Network
Brain networks generated by functional magnetic resonance imaging (fMRI) have shown promising performance in characterizing cerebral social cognition and disorders. However, the scarcity of labeled data has hindered the application of deep graph learning in brain network analysis, accelerating the usage of extra label-free self-supervised contrastive graph learning. However, existing augmentation strategies commonly used in contrastive learning (CL), such as edge and node drop, do not fully benefit brain network learning due to the distribution differences between different modalities of brain neural imaging. To this end, we introduce a novel approach namely spatial–temporal hierarchical augmentation-based contrastive learning (ST-HACL) to enhance the representation learning of functional brain networks. ST-HACL leverages augmentation methods tailored specifically to brain networks. Our method employs an augmentation strategy from both spatial and temporal level during the brain network construction process to generate contrastive samples, enabling label-free self-supervised learning. We evaluate the performance of our approach on the orthostatic hypotension dataset (OH) and the Alzheimer’s disease neuroimaging initiative dataset (ADNI). Results demonstrate that our model surpasses existing graph neural network (GNN) models and graph-CL methods, achieving F1 scores of 80.61% in OH and 73.01% in ADNI. To the best of our knowledge, our study represents the first attempt at applying brain network-specific contrastive augmentation learning to fMRI analysis.
- 24 May 2024
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 45 (10) , 1419-1426
Estimating Flow Direction of Circle of Willis Using Dynamic Arterial Spin-Labeling MR Angiography
The Circle of Willis (COW) is a crucial mechanism for cerebral collateral circulation. This proof-ofconcept study aims to develop and assess an analysis method to characterize the hemodynamics of the arterial segments in COW using arterial spin labeling (ASL) based non-contrast enhanced dynamic magnetic resonance angiography (dMRA). The developed analysis method uses a graph model, bootstrap strategy, and ensemble learning methodologies to determine the time-curve shift from ASL dMRA to estimate the flow direction within the COW. The performance of the method was assessed on 52 subjects, using the flow direction, either antegrade or retrograde, derived from 3D phase contrast (PC) MRI as the reference. A total of 340 arterial segments in COW were evaluated, among which 30 (8.8%) had retrograde flow according to 3D PC. The ASL dMRA-based flow direction estimation has an accuracy, sensitivity, and specificity of 95.47%, 80%, and 96.34%, respectively. Using ASL dMRA and the developed image analysis method to estimate the flow direction in COW is feasible. This study provides a new method to assess the hemodynamics of the COW, which could be useful for the diagnosis and study of cerebrovascular diseases. COW = Circle of Willis; ASL = arterial spin labeling; dMRA =dynamic magnetic resonance angiography; PC = phase contrast.
- 24 May 2024
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 45 (10) , 1419-1426
Estimating Flow Direction of Circle of Willis Using Dynamic Arterial Spin-Labeling MR Angiography
The Circle of Willis (COW) is a crucial mechanism for cerebral collateral circulation. This proof-ofconcept study aims to develop and assess an analysis method to characterize the hemodynamics of the arterial segments in COW using arterial spin labeling (ASL) based non-contrast enhanced dynamic magnetic resonance angiography (dMRA). The developed analysis method uses a graph model, bootstrap strategy, and ensemble learning methodologies to determine the time-curve shift from ASL dMRA to estimate the flow direction within the COW. The performance of the method was assessed on 52 subjects, using the flow direction, either antegrade or retrograde, derived from 3D phase contrast (PC) MRI as the reference. A total of 340 arterial segments in COW were evaluated, among which 30 (8.8%) had retrograde flow according to 3D PC. The ASL dMRA-based flow direction estimation has an accuracy, sensitivity, and specificity of 95.47%, 80%, and 96.34%, respectively. Using ASL dMRA and the developed image analysis method to estimate the flow direction in COW is feasible. This study provides a new method to assess the hemodynamics of the COW, which could be useful for the diagnosis and study of cerebrovascular diseases. COW = Circle of Willis; ASL = arterial spin labeling; dMRA =dynamic magnetic resonance angiography; PC = phase contrast.
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 28 (6) , 3571-3582
Adaptive Knowledge Distillation for High-Quality Unsupervised MRI Reconstruction With Model-Driven Priors
Magnetic Resonance Imaging (MRI) reconstruction has made significant progress with the introduction of Deep Learning (DL) technology combined with Compressed Sensing (CS). However, most existing methods require large fully sampled training datasets to supervise the training process, which may be unavailable in many applications. Current unsupervised models also show limitations in performance or speed and may face unaligned distributions during testing. This paper proposes an unsupervised method to train competitive reconstruction models that can generate high-quality samples in an end-to-end style. Firstly teacher models are trained by filling the re-undersampled images and compared with the undersampled images in a self-supervised manner. The teacher models are then distilled to train another cascade model that can leverage the entire undersampled k-space during its training and testing. Additionally, we propose an adaptive distillation method to re-weight the samples based on the variance of teachers, which represents the confidence of the reconstruction results, to improve the quality of distillation. Experimental results on multiple datasets demonstrate that our method significantly accelerates the inference process while preserving or even improving the performance compared to the teacher model. In our tests, the distilled models show 5%-10% improvements in PSNR and SSIM compared with no distillation and are 10 times faster than the teacher. Source code of our work could be acquired at https://github.com/BITwzl/unsupervised_mri_reconstruction .
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 28 (6) , 3571-3582
Adaptive Knowledge Distillation for High-Quality Unsupervised MRI Reconstruction With Model-Driven Priors
Magnetic Resonance Imaging (MRI) reconstruction has made significant progress with the introduction of Deep Learning (DL) technology combined with Compressed Sensing (CS). However, most existing methods require large fully sampled training datasets to supervise the training process, which may be unavailable in many applications. Current unsupervised models also show limitations in performance or speed and may face unaligned distributions during testing. This paper proposes an unsupervised method to train competitive reconstruction models that can generate high-quality samples in an end-to-end style. Firstly teacher models are trained by filling the re-undersampled images and compared with the undersampled images in a self-supervised manner. The teacher models are then distilled to train another cascade model that can leverage the entire undersampled k-space during its training and testing. Additionally, we propose an adaptive distillation method to re-weight the samples based on the variance of teachers, which represents the confidence of the reconstruction results, to improve the quality of distillation. Experimental results on multiple datasets demonstrate that our method significantly accelerates the inference process while preserving or even improving the performance compared to the teacher model. In our tests, the distilled models show 5%-10% improvements in PSNR and SSIM compared with no distillation and are 10 times faster than the teacher. Source code of our work could be acquired at https://github.com/BITwzl/unsupervised_mri_reconstruction .