Dr. Liang Zhao received the PhD degree in photogrammetry and remote sensing from the Institute of Remote Sensing and Geographic Information System, School of Earth and Space Science, Peking University, Beijing China, in January 2013. From 2012--2014, he was a Postdoctoral Research Fellow in the Centre for Autonomous Systems (CAS), Faculty of Engineering and Information Technology, University of Technology, Sydney (UTS), Australia. From 2014--2016, he worked as a Postdoctoral Research Associate in the Hamlyn Centre for Robotic Surgery, Department of Computing, Faculty of Engineering, Imperial College London, United Kingdom.
He is currently a Lecturer in CAS, UTS since July 2016. His current research interests include surgical robotics, simultaneous localization and mapping (SLAM), monocular SLAM (or Structure-from-Motion), aerial photogrammetry, optimization techniques in mobile robot localization and mapping and image guide robotic surgery.
He serves as reviewer for
- International Journal of Robotics Research (IJRR)
- IEEE Transactions on Intelligent Transportation Systems (T-ITS)
- IEEE Transactions on Automation Science and Engineering (T-ASE)
- ASME Journal of Dynamic Systems, Measurement and Control
- ICRA 2012, ICRA 2013, IROS 2014, ICRA 2015, IROS 2015
- Surgical Robotics
- Simultaneous Localisation and Mapping (SLAM)
- Monocular SLAM (or Structure-from-Motion)
- Aerial Photogrammetry
- Optimisation Techniques in mobile robot localisation and mapping and image guide robotic surgery
- Advanced Robotics (49274)
- Control of Mechatronic Systems (49329)
- Sensors and Control for Mechatronic Systems (41014)
Song, J., Wang, J., Zhao, L., Huang, S. & Dissanayake, G. 2018, 'Dynamic Reconstruction of Deformable Soft-Tissue With Stereo Scope in Minimal Invasive Surgery', IEEE Robotics and Automation Letters, vol. 3, no. 1, pp. 155-162.View/Download from: UTS OPUS or Publisher's site
The classical absolute orientation method is capable of transforming tie points (TPs) from a local coordinate system to a global (geodetic) coordinate system. The method is based only on a unique set of similarity transformation parameters estimated by minimizing the total difference between all ground control points (GCPs) and the fitted points. Nevertheless, it often yields a transformation with poor accuracy, especially in large-scale study cases. To address this problem, this study proposes a novel absolute orientation method based on distance kernel functions, in which various sets of similarity transformation parameters instead of only one set are calculated. When estimating the similarity transformation parameters for TPs using the iterative solution of a non-linear least squares problem, we assigned larger weighting matrices for the GCPs for which the distances from the point are short. The weighting matrices can be evaluated using the distance kernel function as a function of the distances between the GCPs and the TPs. Furthermore, we used the exponential function and the Gaussian function to describe distance kernel functions in this study. To validate and verify the proposed method, six synthetic and two real datasets were tested. The accuracy was significantly improved by the proposed method when compared to the classical method, although a higher computational complexity is experienced
Vander Poorten, E., Zhao, L., Tran, P., Devreker, A., Gruijthuijsen, C., Portoles-Diez, S., Smoljkic, G., Strbac, V., Famaey, N., Reynaerts, D., Vander Sloten, J., Tibebu, A., Yu, B., Rauch, C., Bernard, F., Kassahun, Y., Metzen, J.H., Giannarou, S., Lee, S., Yang, G., Mazomenos, E., Chang, P., Stoyanov, D., Kvasnytsia, M., Van Deun, J., Verhoelst, E., Sette, M., Di Iasio, A., Leo, G., Hertner, F., Scherly, D., Chelini, L., Häni, N., Seatovic, D., Rosa, B., De Praetere, H. & Herijgers, P. 2016, 'Cognitive AutonomouS CAtheters Operating in Dynamic Environments', Journal of Medical Robotics Research, vol. 1, no. 3.View/Download from: UTS OPUS or Publisher's site
Advances in miniaturized surgical instrumentation are key to less demanding and safer medical interventions. In cardiovascular
procedures interventionalists turn towards catheter-based interventions, treating patients considered unfit for more invasive approaches.
A positive outcome is not guaranteed. The risk for calcium dislodgement, tissue damage or even vessel rupture cannot
be eliminated when instruments are maneuvered through fragile and diseased vessels. This paper reports on the progress made in
terms of catheter design, vessel reconstruction, catheter shape modeling, surgical skill analysis, decision-making and control. These
efforts are geared towards the development of the necessary technology to autonomously steer catheters through the vasculature,
a target of the EU-funded project CASCADE (Cognitive AutonomouS CAtheters operating in Dynamic Environments). Whereas
autonomous placement of an aortic valve implant forms the ultimate and concrete goal, the technology of individual building blocks
to reach such ambitious goal is expected to be much sooner impacting and assisting interventionalists in their daily clinical practice.
Zhao, L., Giannarou, S., Lee, S.-.L. & Yang, G.Z. 2016, 'SCEM+: Real-time Robust Simultaneous Catheter and Environment Modelling for Endo-vascular Navigation', IEEE Robotics and Automation Letters, vol. 1, no. 2, pp. 961-968.View/Download from: UTS OPUS or Publisher's site
Endovascular procedures are characterised by significant challenges mainly due to the complexity in catheter control and navigation. Real-time recovery of the 3-D structure of the vasculature is necessary to visualise the interaction between the catheter and its surrounding environment to facilitate catheter manipulations. State-of-the-art intraoperative vessel reconstruction approaches are increasingly relying on nonionising imaging techniques such as optical coherence tomography (OCT) and intravascular ultrasound (IVUS). To enable accurate recovery of vessel structures and to deal with sensing errors and abrupt catheter motions, this letter presents a robust and real-time vessel reconstruction scheme for endovascular navigation based on IVUS and electromagnetic (EM) tracking. It is formulated as a nonlinear optimisation problem, which considers the uncertainty in both the IVUS contour and the EM pose, as well as vessel morphology provided by preoperative data. Detailed phantom validation is performed and the results demonstrate the potential clinical value of the technique.
Sun, Y., Zhao, L., Huang, S., Yan, L. & Dissanayake, G. 2015, 'Line matching based on planar homography for stereo aerial images', ISPRS Journal of Photogrammetry and Remote Sensing, vol. 104, pp. 1-17.View/Download from: UTS OPUS or Publisher's site
© 2015 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). We propose an efficient line matching algorithm for a pair of calibrated aerial photogrammetric images, which makes use of sparse 3D points triangulated from 2D point feature correspondences to guide line matching based on planar homography. Two different strategies are applied in the proposed line matching algorithm for two different cases. When three or more points can be found coplanar with the line segment to be matched, the points are used to fit a plane and obtain an accurate planar homography. When one or two points can be found, the approximate terrain plane parallel to the line segment is utilized to compute an approximate planar homography. Six pairs of rural or urban aerial images are used to demonstrate the efficiency and validity of the proposed algorithm. Compared with line matching based on 2D point feature correspondences, the proposed method can increase the number of correctly matched line segments. In addition, compared with most line matching methods that do not use 2D point feature correspondences, the proposed method has better efficiency, although it obtains fewer matches. The C/C++ source code for the proposed algorithm is available at http://services.eng.uts.edu.au/~sdhuang/research.htm.
Zhao, L., Huang, S., Sun, Y., Yan, L. & Dissanayake, G. 2015, 'ParallaxBA: Bundle adjustment using parallax angle feature parametrization', International Journal of Robotics Research, vol. 34, no. 4-5, pp. 493-516.View/Download from: UTS OPUS or Publisher's site
©The Author(s) 2015. The main contribution of this paper is a novel feature parametrization based on parallax angles for bundle adjustment (BA) in structure and motion estimation from monocular images. It is demonstrated that under certain conditions, describing feature locations using their Euclidean XYZ coordinates or using inverse depth in BA leads to ill-conditioned normal equations as well as objective functions that have very small gradients with respect to some of the parameters describing feature locations. The proposed parallax angle feature parametrization in BA (ParallaxBA) avoids both of the above problems leading to better convergence properties and more accurate motion and structure estimates. Simulation and experimental datasets are used to demonstrate the impact of different feature parametrizations on BA, and the improved convergence, efficiency and accuracy of the proposed ParallaxBA algorithm when compared with some existing BA packages such as SBA, sSBA and g2o. The C/C++ source code of ParallaxBA is available on OpenSLAM (https://openslam.org/).
Zhao, L., Huang, S., Yan, L. & Dissanayake, G. 2015, 'A new feature parametrization for monocular SLAM using line features', Robotica, vol. 33, no. 3, pp. 513-536.View/Download from: UTS OPUS or Publisher's site
© 2014 Cambridge University Press. This paper presents a new monocular SLAM algorithm that uses straight lines extracted from images to represent the environment. A line is parametrized by two pairs of azimuth and elevation angles together with the two corresponding camera centres as anchors making the feature initialization relatively straightforward. There is no redundancy in the state vector as this is a minimal representation. A bundle adjustment (BA) algorithm that minimizes the reprojection error of the line features is developed for solving the monocular SLAM problem with only line features. A new map joining algorithm which can automatically optimize the relative scales of the local maps is used to combine the local maps generated using BA. Results from both simulations and experimental datasets are used to demonstrate the accuracy and consistency of the proposed BA and map joining algorithms.
Sun, Y., Zhao, L., Huang, S., Yan, L. & Dissanayake, G. 2014, 'L2-SIFT: SIFT feature extraction and matching for large images in large-scale aerial photogrammetry', ISPRS Journal of Photogrammetry and Remote Sensing, vol. 91, pp. 1-16.View/Download from: UTS OPUS or Publisher's site
The primary contribution of this paper is an efficient feature extraction and matching implementation for large images in large-scale aerial photogrammetry experiments. First, a Block-SIFT method is designed to overcome the memory limitation of SIFT for extracting and matching features from large photogrammetric images. For each pair of images, the original large image is split into blocks and the possible corresponding blocks in the other image are determined by pre-estimating the relative transformation between the two images. Because of the reduced memory requirement, eatures can be extracted and matched from the original images without down-sampling. Next, a red-black tree data structure is applied to create a feature relationship to reduce the search complexity when matching tie points. Meanwhile, tree key exchange and segment matching methods are proposed to match the tie points along-track and across-track. Finally, to evaluate the accuracy of the features extracted and matched from the proposed L2-SIFT algorithm, a bundle adjustment with parallax angle feature parametrization (ParallaxBA) is applied to obtain the Mean Square Error (MSE) of the feature reprojections, where the feature extraction and matching result is the only information used in the nonlinear optimisation system. Seven different experimental aerial photogrammetric datasets are used to demonstrate the efficiency and validity of the proposed algorithm. It is demonstrated that more than 33 million features can be extracted and matched from the Taian dataset with 737 images within 21h using the L2-SIFT algorithm. In addition, the ParallaxBA involving more than 2.7 million features and 6 million image points can easily converge to an MSE of 0.03874. The C/C++ source code for the proposed algorithm is available at http://services.eng.uts.edu.au/~sdhuang/research.htm.
Huang, S., Wang, J., Zhao, L., Ge, X., Zhang, C., He, X. & Wang, X. 2017, 'High Quality 3D Reconstruction of Indoor Environments using RGB-D Sensors', IEEE 12th Conference on Industrial Electronics and Applications (ICIEA), Siem Reap, Cambodia, pp. 1736-1741.View/Download from: UTS OPUS
Song, J., wang, J., Zhao, L., Huang, S. & Dissanayake, G. 2017, 'Deformable Soft-tissue Reconstruction using Stereo Scope for Minimal Invasive Surgery', CARS 2017 -- Computer Assisted Radiology and Surgery, 31st International Congress and Exhibition, CARS 2017 -- Computer Assisted Radiology and Surgery, 31st International Congress and Exhibition.View/Download from: UTS OPUS
song, J., wang, J., Zhao, L., huang, S. & Dissanayake, G. 2017, 'Robust Shape Recovery of Deformable Soft-tissue Based on Information from Stereo Scope for Minimal Invasive Surgery', Hamlyn Symposium on Medical Robotics (HSMR 2017), Hamlyn Symposium on Medical Robotics (HSMR 2017).View/Download from: UTS OPUS
wang, J., song, J., Zhao, L. & huang, S. 2017, 'A Submap Joining Based RGB-D SLAM Algorithm using Planes as Features', 11th Conference on Field and Service Robotics (FSR 2017), 11th Conference on Field and Service Robotics (FSR 2017), Springer, Zurich, Switzerland.View/Download from: UTS OPUS
Liu, L., Wang, Y., Zhao, L. & Huang, S. 2017, 'Evaluation of Different SLAM Algorithms usingGoogle Tangle Data', IEEE Conference on Industrial Electronics and Applications, Siem Reap, Cambodia..View/Download from: UTS OPUS
Zhao, L., Giannarou, S., Lee, S.-.L. & Yang, G.Z. 2016, 'Registration-free Simultaneous Catheter and Environment Modelling', Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016 (LNCS), Medical Image Computing and Computer-Assisted Intervention, Springer, Athens, Greece, pp. 525-533.View/Download from: UTS OPUS or Publisher's site
Endovascular procedures are challenging to perform due to the complexity and difficulty in catheter manipulation. The simultaneous recovery of the 3D structure of the vasculature and the catheter position and orientation intra-operatively is necessary in catheter control and navigation. State-of-art Simultaneous Catheter and Environment Modelling provides robust and real-time 3D vessel reconstruction based on real-time intravascular ultrasound (IVUS) imaging and electromagnetic (EM) sensing, but still relies on accurate registration between EM and pre-operative data. In this paper, a registration-free vessel reconstruction method is proposed for endovascular navigation. In the optimisation framework, the EM-CT registration is estimated and updated intra-operatively together with the 3D vessel reconstruction from IVUS, EM and pre-operative data, and thus does not require explicit registration. The proposed algorithm can also deal with global (patient) motion and periodic deformation caused by cardiac motion. Phantom and in-vivo experiments validate the accuracy of the algorithm and the results demonstrate the potential clinical value of the technique.
Zhao, L., Giannarou, S., Lee, S.-.L., Merrifield, R. & Yang, G.Z. 2016, 'Intra-operative Simultaneous Catheter and Environment Modelling for Endovascular Navigation Based on Intravascular Ultrasound, Electromagnetic Tracking and Pre-operative Data', Proceedings of The Hamlyn Symposium on Medical Robotics, The Hamlyn Symposium on Medical Robotics, Imperial College London and the Royal Geographical Society, London, UK, pp. 76-77.View/Download from: UTS OPUS
Cardiovascular diseases (CVD) form the single most
common cause of death. Catheter procedures are among
the most common surgical interventions used to treat
CVD. Due to their minimal access trauma, these
procedures extend the range of patients able to receive
interventional CVD treatment to age groups dominated
by co-morbidity and unacceptable risks for open surgery
. The downside associated with minimising access
incisions lies at the increased complexity and difficult
manipulation of the instruments and anatomical targets,
which is mainly caused by the loss of direct access to
the anatomy and the poor visualisation of the surgical
site. The current clinical approaches to endovascular
procedures mainly rely on 2D guidance based on X-ray
fluoroscopy, which uses ionising radiation and
dangerous contrast agents .
In this paper, a Simultaneous Catheter and Environment
Modelling (SCEM) method is presented for
endovascular navigation based on intravascular
ultrasound (IVUS) imaging, electromagnetic (EM)
sensing as well as the vessel structure information
provided from the pre-operative CT/MR imaging (see
Fig. 1). Thus, radiation dose and contrast agents are
avoided. The proposed SCEM intra-operatively recovers
the 3D structure of the vasculature together with the
pose of the catheter tip, which the knowledge of the
interaction between the catheter and its surroundings
can be provided. The corresponding uncertainties of
both vessel reconstruction and catheter pose can also be
computed which is necessary for autonomous robotic
catheter navigation. Experimental results using three
different phantoms, with different catheter motions and
cardiac motions simulated by using a periodic pump
demonstrated the accuracy of the vessel reconstruction
and the potential clinical value of the proposed SCEM
Zhao, L., Huang, S. & Dissanayake, G. 2014, 'Linear MonoSLAM: A Linear Approach to Large-Scale Monocular SLAM Problems', 2014 IEEE International Conference on Robotics & Automation (ICRA), IEEE International Conference on Robotics and Automation, IEEE, Hong Kong, China, pp. 1517-1523.View/Download from: UTS OPUS or Publisher's site
This paper presents a linear approach for solving monocular simultaneous localization and mapping (SLAM) problems. The algorithm rst builds a sequence of small initial submaps and then joins these submaps together in a divideand-conquer (D&C) manner. Each of the initial submap is built using three monocular images by bundle adjustment (BA), which is a simple nonlinear optimization problem. Each step in the D&C submap joining is solved by a linear least squares together with a coordinate and scale transformation. Since the only nonlinear part is in the building of the initial submaps, the algorithm makes it possible to solve large-scale monocular SLAM while avoiding issues associated with initialization, iteration, and local minima that are present in most of the nonlinear optimization based algorithms currently used for large-scale monocular SLAM. Experimental results based on publically available datasets are used to demonstrate that the proposed algorithms yields solutions that are very close to those obtained using global BA starting from good initial guess.
Zhao, L., Huang, S. & Dissanayake, G. 2013, 'Linear SLAM: A Linear Solution to the Feature-based and Pose Graph SLAM based on Submap Joining', 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Tokyo, Japan, pp. 24-30.View/Download from: UTS OPUS or Publisher's site
This paper presents a strategy for large-scale SLAM through solving a sequence of linear least squares problems. The algorithm is based on submap joining where submaps are built using any existing SLAM technique. It is demonstrated that if submaps coordinate frames are judiciously selected, the least squares objective function for joining two submaps becomes a quadratic function of the state vector. Therefore, a linear solution to large-scale SLAM that requires joining a number of local submaps either sequentially or in a more efficient Divide and Conquer manner, can be obtained. The proposed Linear SLAM technique is applicable to both feature-based and pose graph SLAM, in two and three dimensions, and does not require any assumption on the character of the covariance matrices or an initial guess of the state vector. Although this algorithm is an approximation to the optimal full nonlinear least squares SLAM, simulations and experiments using publicly available datasets in 2D and 3D show that Linear SLAM produces results that are very close to the best solutions that can be obtained using full nonlinear optimization started from an accurate initial value. The C/C++ and MATLAB source codes for the proposed algorithm are available on OpenSLAM.
Ahmad, A., Zhao, L., Huang, S. & Dissanayake, G. 2012, 'Convergence comparison of least squares based bearing-only SLAM algorithms using different landmark parametrizations', 2012 International Conference on Control, Automation, Robotics & Vision (ICARCV ), International Conference on Control, Automation, Robotics and Vision, IEEE, Guangzhou, China, pp. 1006-1011.View/Download from: UTS OPUS or Publisher's site
This paper compares the convergence of least squares based 2D bearing-only SLAM algorithms using different landmark parametrizations. It is shown that the requirement on the accuracy of the initial value vary significantly when using different landmark parametrizations. Especially, for small scale bearing-only SLAM problems, the region of attraction of the global minimum for Gauss-Newton iteration based bearing-only SLAM algorithm using parallax angle landmark parametrization is significantly larger as compared with those of bearing-only SLAM algorithms using other landmark parametrizations.
Himstedt, M., Alempijevic, A., Zhao, L., Huang, S. & Boehme, H. 2012, 'Towards robust vision-based self-localization of vehicles in dense urban environments', 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Algarve, Portugal, pp. 3152-3157.View/Download from: UTS OPUS or Publisher's site
Self-localization of ground vehicles in densely populated urban environments poses a significant challenge. The presence of tall buildings in close proximity to traversable areas limits the use of GPS-based positioning techniques in such environments. This paper presents an approach to global localization on a hybrid metric-topological map using a monocular camera and wheel odometry. The global topology is built upon spatially separated reference places represented by local image features. In contrast to other approaches we employ a feature selection scheme ensuring a more discriminative representation of reference places while simultaneously rejecting a multitude of features caused by dynamic objects. Through fusion with additional local cues the reference places are assigned discrete map positions allowing metric localization within the map. The self-localization is carried out by associating observed visual features with those stored for each reference place. Comprehensive experiments in a dense urban environment covering a time difference of about 9 months are carried out. This demonstrates the robustness of our approach in environments subjected to high dynamic and environmental changes.
Hu, G., Huang, S., Zhao, L., Alempijevic, A. & Dissanayake, G. 2012, 'A Robust RGB-D SLAM algorithm', 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Algarve, Portugal, pp. 1174-1179.View/Download from: UTS OPUS or Publisher's site
Recently RGB-D sensors have become very popular in the area of Simultaneous Localisation and Mapping (SLAM). The major advantage of these sensors is that they provide a rich source of 3D information at relatively low cost. Unfortunately, these sensors in their current forms only have a range accuracy of up to 4 metres. Many techniques which perform SLAM using RGB-D cameras rely heavily on the depth and are restrained to office type and geometrically structured environments. In this paper, a switching based algorithm is proposed to heuristically choose between RGB-BA and RGBD-BA based local maps building. Furthermore, a low cost and consistent optimisation approach is used to join these maps. Thus the potential of both RGB and depth image information are exploited to perform robust SLAM in more general indoor cases. Validation of the proposed algorithm is performed by mapping a large scale indoor scene where traditional RGB-D mapping techniques are not possible.
Zhao, L., Huang, S., Yan, L. & Dissanayake, G. 2011, 'Parallax angle parametrization for monocular SLAM', 2011 IEEE International Conference on Robotics and Automation (ICRA), IEEE International Conference on Robotics and Automation, IEEE, Shanghai, China, pp. 3117-3124.View/Download from: UTS OPUS or Publisher's site
This paper presents a new unified feature parametrization approach for monocular SLAM. The parametrization is based on the parallax angle and can reliably represent both nearby and distant features, as well as features in the direction of camera motion and features observed only once. A new bundle adjustment (BA) algorithm using the proposed parallax angle parametrization is developed and shown to be more reliable as compared with existing BA algorithms that use Euclidean XYZ or inverse depth parametrizations. A new map joining algorithm that allows combining a sequence of local maps generated using BA with the proposed parametrization, that avoids the large computational cost of a global BA, and can automatically optimize the relative scales of the local maps without any loss of information, is also presented. Extensive simulations and a publicly available large-scale real dataset with centimeter accuracy ground truth are used to demonstrate the accuracy and consistency of the BA and map joining algorithms using the new parametrization. Especially, since the relative scales are optimized automatically in the proposed BA and map joining algorithms, there is no need to compute any relative scales even for a loop more than 1km.
Zhao, L., Huang, S., Yan, L., Wang, J., Hu, G. & Dissanayake, G. 2010, 'Large-Scale Monocular SLAM by Local Bundle Adjustment and Map Joining', 2010 International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Singapore, pp. 431-436.View/Download from: UTS OPUS or Publisher's site
This paper first demonstrates an interesting property of bundle adjustment (BA), âscale drift correctionâ. Here âscale drift correctionâ means that BA can converge to the correct solution (up to a scale) even if the initial values of the camera pose translations and point feature positions are calculated using very different scale factors. This property together with other properties of BA makes it the best approach for monocular Simultaneous Localization and Mapping (SLAM), without considering the computational complexity. This naturally leads to the idea of using local BA and map joining to solve large-scale monocular SLAM problem, which is proposed in this paper. The local maps are built through Scale-Invariant Feature Transform (SIFT) for feature detection and matching, random sample consensus (RANSAC) paradigm at different levels for robust outlier removal, and BA for optimization. To reduce the computational cost of the large-scale map building, the features in each local map are judiciously selected and then the local maps are combined using a recently developed 3D map joining algorithm. The proposed large-scale monocular SLAM algorithm is evaluated using a publicly available dataset with centimeter-level ground truth.
Sun, Y. & Zhao, L. 2015, 'Line Matching based on Planar Homography for Aerial Photogrammetric Images C/C++ source code'.
C/C++ source code
C/C++ source code
Sun, Y. & Zhao, L. 2013, 'L2-SIFT: SIFT Feature Extraction and Matching for Large Images in Large-scale Aerial Photogrammetry C/C++ source code'.
C/C++ source code
MATLAB and C/C++ source code
- Imperial College London
- University of Leeds
- Peking University
- Zhejiang University
- Tianjin University
- Northeast University