Xuesong Gao
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
Xuesong Gao has a total of 62 co-authors. Here are the top co-authors by number of shared publications:
Bin Wei
1 shared publicationBingqiang Wang
Shandong WeiGao Surgical Robot Co. Ltd. Weihai China
1 shared publicationChen Wang
The Affiliated Hospital of Qingdao University Qingdao China
1 shared publication- Chi-Hung Chi
Digital Trust Centre, Nanyang Technological University, Singapore
1 shared publication - Dong Zhang
College of Horticulture, Yangling Sub-Center of National Center for Apple Improvement, Northwest A&F University , Yangling 712100, Shaanxi , P. R. China
1 shared publication Fengju Guan
1 shared publication
Latest Publications
- 9 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 8940-8953
Multi-Layer Transfer Learning for Cross-Domain Recommendation Based on Graph Node Representation Enhancement
Effectively representing and transferring user preferences across various domains presents a significant challenge in cross-domain recommendation (CDR). Some approaches utilize graph neural networks that use interaction behavior to establish relationships between entities, providing a comprehensive understanding of user interests. However, the impact of consistent semantics across various types, fields, and perspectives of social media information on user preferences is overlooked, i.e. the multidimensional consistency of user preferences. This oversight results in graph node representations that inadequately reflect user preferences. To address these limitations, we propose a multi-layer transfer learning network (MTLG) for CDR based on graph node representation enhancement via multi-dimensional consistent user preferences. Firstly, the model introduces a set of globally shared semantic units to perform different-grained semantic alignment of multiple media information without clear alignment boundaries, thereby modeling multi-dimensional consistent user preference features. These features are then seamlessly integrated with the initial high-order graph structure embedding features, thus significantly improving the quality of graph node representation. Secondly, the model innovatively designs a multi-layer transfer learning network that hierarchically aligns the domain distribution differences. It calculates the similarity between domains to derive layer weights for more precise transfer learning, thereby mitigating the possibility of information error accumulation resulting from inaccurate feature aggregation processes. We conducted numerous experiments on 3 scenarios, including 7,954,943 rating information from the Amazon dataset. The results indicate that MTLG's recommendation accuracy surpasses those of state-of-the-art methods.
- 9 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 8940-8953
Multi-Layer Transfer Learning for Cross-Domain Recommendation Based on Graph Node Representation Enhancement
Effectively representing and transferring user preferences across various domains presents a significant challenge in cross-domain recommendation (CDR). Some approaches utilize graph neural networks that use interaction behavior to establish relationships between entities, providing a comprehensive understanding of user interests. However, the impact of consistent semantics across various types, fields, and perspectives of social media information on user preferences is overlooked, i.e. the multidimensional consistency of user preferences. This oversight results in graph node representations that inadequately reflect user preferences. To address these limitations, we propose a multi-layer transfer learning network (MTLG) for CDR based on graph node representation enhancement via multi-dimensional consistent user preferences. Firstly, the model introduces a set of globally shared semantic units to perform different-grained semantic alignment of multiple media information without clear alignment boundaries, thereby modeling multi-dimensional consistent user preference features. These features are then seamlessly integrated with the initial high-order graph structure embedding features, thus significantly improving the quality of graph node representation. Secondly, the model innovatively designs a multi-layer transfer learning network that hierarchically aligns the domain distribution differences. It calculates the similarity between domains to derive layer weights for more precise transfer learning, thereby mitigating the possibility of information error accumulation resulting from inaccurate feature aggregation processes. We conducted numerous experiments on 3 scenarios, including 7,954,943 rating information from the Amazon dataset. The results indicate that MTLG's recommendation accuracy surpasses those of state-of-the-art methods.
- Open Access
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 11222-11225
Generalized Sampling of Non-Local Textural Clues Multi-View Stereo Framework
3D reconstruction remains a pivotal topic in contemporary multimedia research, with multi-view stereo (MVS) methods being instrumental. Traditional MVS algorithms, have demonstrated robust performance in large-scale scenes. However, these methods often struggle with depth estimation in non-textured regions, resulting in inconsistencies across different areas of the reconstructed 3D model. This paper introduces a novel MVS framework, General Sampling Non-local Texture Features (GST-MVS), designed to enhance depth estimation by leveraging the intrinsic relationship between image texture and object depth. Our approach optimizes the 3D reconstruction process, yielding superior results, especially in low-textured regions. Comprehensive evaluations on the ETH3D dataset indicate that GST-MVS achieves dense and accurate 3D point cloud reconstructions, outperforming existing MVS algorithms. The GST-MVS framework is open-source and accessible on GitHub at https://github.com/Jasmine-tjy/GST-mvs. Additionally, Docker image facilitating the required execution environment, which is available at https://hub.docker.com/r/tangjas111/cmake_cuda_opencv/tags.
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 11222-11225
Open AccessGeneralized Sampling of Non-Local Textural Clues Multi-View Stereo Framework
3D reconstruction remains a pivotal topic in contemporary multimedia research, with multi-view stereo (MVS) methods being instrumental. Traditional MVS algorithms, have demonstrated robust performance in large-scale scenes. However, these methods often struggle with depth estimation in non-textured regions, resulting in inconsistencies across different areas of the reconstructed 3D model. This paper introduces a novel MVS framework, General Sampling Non-local Texture Features (GST-MVS), designed to enhance depth estimation by leveraging the intrinsic relationship between image texture and object depth. Our approach optimizes the 3D reconstruction process, yielding superior results, especially in low-textured regions. Comprehensive evaluations on the ETH3D dataset indicate that GST-MVS achieves dense and accurate 3D point cloud reconstructions, outperforming existing MVS algorithms. The GST-MVS framework is open-source and accessible on GitHub at https://github.com/Jasmine-tjy/GST-mvs. Additionally, Docker image facilitating the required execution environment, which is available at https://hub.docker.com/r/tangjas111/cmake_cuda_opencv/tags.
- 8 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 71 (2) , 6627-6641
Privacy-Preserving 3-D Skeleton-Based Video Action Recognition via Graph Convolution Network
In recent years, 3D skeleton data-based human action recognition has attracted an increasing number of researchers because of its high robustness under illumination change and scene variation. However, in 3D skeleton data-based human action recognition, there are still some challenges including multi-object recognition and private data leakage. To address this issue, in this paper, we propose a privacy-preserving end-to-end action recognition model based on a Graph Convolutional Network (GCN). Specifically, we innovatively propose a 3D pose estimation method to obtain 3D human skeleton data from color data streams and depth data streams, in which the mean filtering technique is applied to protect users’ video data privacy while not affecting the extraction of the skeleton data. Then, we propose a feature representation method and represent the skeleton data to a GCN model, which convolves the centroid of the human skeleton, four limbs, and the trunk of a person. After that, we propose a method to fuse and extract multiple persons’ movement features and propose a novel GCN-based interactive action recognition model to recognize multiple people’s skeleton data. Moreover, to protect recognition model privacy, we also design a ciphertext-based secure action recognition method, which guarantees the confidentiality of model parameters during action recognition. Finally, we evaluate the performance of our model in real datasets, and the results demonstrate that our model can achieve high recognition accuracy than existing models. Meanwhile, experimental results demonstrate that the mean filtering technique can well protect the privacy of the appearance data and the information of skeleton data can be well preserved.
- 8 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 71 (2) , 6627-6641
Privacy-Preserving 3-D Skeleton-Based Video Action Recognition via Graph Convolution Network
In recent years, 3D skeleton data-based human action recognition has attracted an increasing number of researchers because of its high robustness under illumination change and scene variation. However, in 3D skeleton data-based human action recognition, there are still some challenges including multi-object recognition and private data leakage. To address this issue, in this paper, we propose a privacy-preserving end-to-end action recognition model based on a Graph Convolutional Network (GCN). Specifically, we innovatively propose a 3D pose estimation method to obtain 3D human skeleton data from color data streams and depth data streams, in which the mean filtering technique is applied to protect users’ video data privacy while not affecting the extraction of the skeleton data. Then, we propose a feature representation method and represent the skeleton data to a GCN model, which convolves the centroid of the human skeleton, four limbs, and the trunk of a person. After that, we propose a method to fuse and extract multiple persons’ movement features and propose a novel GCN-based interactive action recognition model to recognize multiple people’s skeleton data. Moreover, to protect recognition model privacy, we also design a ciphertext-based secure action recognition method, which guarantees the confidentiality of model parameters during action recognition. Finally, we evaluate the performance of our model in real datasets, and the results demonstrate that our model can achieve high recognition accuracy than existing models. Meanwhile, experimental results demonstrate that the mean filtering technique can well protect the privacy of the appearance data and the information of skeleton data can be well preserved.
- 4 May 2022
- Published by Wiley in The International Journal of Medical Robotics and Computer Assisted Surgery
- Vol. 18 (4)
Application of 5G technology to conduct tele‐surgical robot‐assisted laparoscopic radical cystectomy
Background: The aim of this study was to test the effectiveness, safety and stability of the 5G communication technology in clinical laparoscopic telesurgery.Methods: An ultra‐remote radical cystectomy (network communication distance of nearly 3000 km) was performed on patient diagnosed with T2N0M0 stage bladder cancer using a domestically produced “MicroHand” surgical robot.Results: The network delay, operative time, blood loss, intraoperative complications, postoperative recovery, and hospitalisation time were recorded. The 5G network was used throughout the operation, with an average total delay of 254 ms. The operation went well and the patient recovered smoothly.Conclusions: Ultra‐remote clinical laparoscopic surgery can be performed safely and smoothly. More importantly, our model can provide insights for promoting the future development of telesurgery in China.
- 4 May 2022
- Published by Wiley in The International Journal of Medical Robotics and Computer Assisted Surgery
- Vol. 18 (4)
Application of 5G technology to conduct tele‐surgical robot‐assisted laparoscopic radical cystectomy
Background: The aim of this study was to test the effectiveness, safety and stability of the 5G communication technology in clinical laparoscopic telesurgery.Methods: An ultra‐remote radical cystectomy (network communication distance of nearly 3000 km) was performed on patient diagnosed with T2N0M0 stage bladder cancer using a domestically produced “MicroHand” surgical robot.Results: The network delay, operative time, blood loss, intraoperative complications, postoperative recovery, and hospitalisation time were recorded. The 5G network was used throughout the operation, with an average total delay of 254 ms. The operation went well and the patient recovered smoothly.Conclusions: Ultra‐remote clinical laparoscopic surgery can be performed safely and smoothly. More importantly, our model can provide insights for promoting the future development of telesurgery in China.
- Open Access
High-Resolution Image Harmonization via Collaborative Dual Transformations
Given a composite image, image harmonization aims to adjust the foreground to make it compatible with the background. High-resolution image harmonization is in high demand, but still remains unexplored. Conventional image harmonization methods learn global RGB-to-RGB transformation which could effortlessly scale to high resolution, but ignore diverse local context. Recent deep learning methods learn the dense pixel-to-pixel transformation which could generate harmonious outputs, but are highly constrained in low resolution. In this work, we propose a high-resolution image harmonization network with Collaborative Dual Transformation (CDTNet) to combine pixel-to-pixel transformation and RGB-to-RGB transformation coherently in an end-to-end network. Our CDTNet consists of a low-resolution generator for pixel-to-pixel transformation, a color mapping module for RGB-to-RGB transformation, and a refinement module to take advantage of both. Extensive experiments on high-resolution benchmark dataset and our created high-resolution real composite images demonstrate that our CDTNet strikes a good balance between efficiency and effectiveness. Our used datasets can be found in https://github.com/bcmi/CDTNet-High-Resolution-Image-Harmonization.
Open AccessHigh-Resolution Image Harmonization via Collaborative Dual Transformations
Given a composite image, image harmonization aims to adjust the foreground to make it compatible with the background. High-resolution image harmonization is in high demand, but still remains unexplored. Conventional image harmonization methods learn global RGB-to-RGB transformation which could effortlessly scale to high resolution, but ignore diverse local context. Recent deep learning methods learn the dense pixel-to-pixel transformation which could generate harmonious outputs, but are highly constrained in low resolution. In this work, we propose a high-resolution image harmonization network with Collaborative Dual Transformation (CDTNet) to combine pixel-to-pixel transformation and RGB-to-RGB transformation coherently in an end-to-end network. Our CDTNet consists of a low-resolution generator for pixel-to-pixel transformation, a color mapping module for RGB-to-RGB transformation, and a refinement module to take advantage of both. Extensive experiments on high-resolution benchmark dataset and our created high-resolution real composite images demonstrate that our CDTNet strikes a good balance between efficiency and effectiveness. Our used datasets can be found in https://github.com/bcmi/CDTNet-High-Resolution-Image-Harmonization.
- Open Access
- 25 October 2021
- Published by Wiley in Security and Communication Networks
- Vol. 2021, 1-15
An Efficient and Privacy-Preserving Biometric Identification Scheme Based on the FITing-Tree
Biometric identification services have been applied to almost all aspects of life. However, how to securely and efficiently identify an individual in a huge biometric dataset is still very challenging. For one thing, biometric data is very sensitive and should be kept secure during the process of biometric identification. On the other hand, searching a biometric template in a large dataset can be very time-consuming, especially when some privacy-preserving measures are adopted. To address this problem, we propose an efficient and privacy-preserving biometric identification scheme based on the FITing-tree, iDistance, and a symmetric homomorphic encryption (SHE) scheme with two cloud servers. With our proposed scheme, the privacy of the user’s identification request and service provider’s dataset is guaranteed, while the computational costs of the cloud servers in searching the biometric dataset can be kept at an acceptable level. Detailed security analysis shows that the privacy of both the biometric dataset and biometric identification request is well protected during the identification service. In addition, we implement our proposed scheme and compare it to a previously reported M-Tree based privacy-preserving identification scheme in terms of computational and communication costs. Experimental results demonstrate that our proposed scheme is indeed efficient in terms of computational and communication costs while identifying a biometric template in a large dataset.
- 25 October 2021
- Published by Wiley in Security and Communication Networks
- Vol. 2021, 1-15
Open AccessAn Efficient and Privacy-Preserving Biometric Identification Scheme Based on the FITing-Tree
Biometric identification services have been applied to almost all aspects of life. However, how to securely and efficiently identify an individual in a huge biometric dataset is still very challenging. For one thing, biometric data is very sensitive and should be kept secure during the process of biometric identification. On the other hand, searching a biometric template in a large dataset can be very time-consuming, especially when some privacy-preserving measures are adopted. To address this problem, we propose an efficient and privacy-preserving biometric identification scheme based on the FITing-tree, iDistance, and a symmetric homomorphic encryption (SHE) scheme with two cloud servers. With our proposed scheme, the privacy of the user’s identification request and service provider’s dataset is guaranteed, while the computational costs of the cloud servers in searching the biometric dataset can be kept at an acceptable level. Detailed security analysis shows that the privacy of both the biometric dataset and biometric identification request is well protected during the identification service. In addition, we implement our proposed scheme and compare it to a previously reported M-Tree based privacy-preserving identification scheme in terms of computational and communication costs. Experimental results demonstrate that our proposed scheme is indeed efficient in terms of computational and communication costs while identifying a biometric template in a large dataset.
- 2 July 2017
- Published by Springer Nature
- p. 3-18
On the Future of Solution Composition in Software Ecosystems
- 2 July 2017
- Published by Springer Nature
- p. 3-18
On the Future of Solution Composition in Software Ecosystems