Juan Zhao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Juan Zhao has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Yuping Ye
School of Smart Marine Science and Technology and Fujian Provincial Key Laboratory of Marine Smart Equipment, Fujian University of Technology, Fuzhou, China
4 shared publications - Zhan Song
Chinese Academy of Sciences
4 shared publications - Jixin Liang
Department of Electronic Information Engineering, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
2 shared publications - Juncheng Han
Chinese Academy of Sciences, Shenzhen Institute of Advanced Technology, Shenzhen, China
2 shared publications - Chuan Xu
University of Science and Technology of China, China and Shenzhen Institute of AdvancedTechnology, Chinese Academy of Sciences, China
1 shared publication - Feifei Gu
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences
1 shared publication
Latest Publications
- 26 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (7) , 9212-9224
CF-GAT: Curvature-Fused Graph Attention Network for High-Precision Unordered Facial Point Cloud Landmark Detection
Due to the lack of large-scale, accurately annotated 3D facial datasets, most current 3D facial landmark detection algorithms rely on 2D texture assistance or non-real digital 3D faces. The performance of these algorithms is limited by the accuracy of texture mapping and the difference between digital faces and actual faces. To tackle these challenges, we built a large-scale, high-precision, and accurately annotated 3D facial database using a structured light system. Building upon this foundation, we proposed a novel point cloud sampling method and 3D facial landmark detection algorithm. This method utilizes a curvature-fused graph attention network to directly predict landmark coordinates from 3D point clouds. Firstly, we extracted a simplifying point set carrying curvature information from the original 3D facial point cloud via geometric point sampling. Then, we incorporated curvature-encoded positional information as the learning component of the attention module and employed it as a feature extractor to construct the network. We evaluated the performance of the method on BU-3DFE, FaceScape and our custom dataset. Compared to existing facial landmark detection algorithms, our model achieved higher accuracy. To facilitate future research on face related applications, we have made the database available on Github 1.
- 26 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (7) , 9212-9224
CF-GAT: Curvature-Fused Graph Attention Network for High-Precision Unordered Facial Point Cloud Landmark Detection
Due to the lack of large-scale, accurately annotated 3D facial datasets, most current 3D facial landmark detection algorithms rely on 2D texture assistance or non-real digital 3D faces. The performance of these algorithms is limited by the accuracy of texture mapping and the difference between digital faces and actual faces. To tackle these challenges, we built a large-scale, high-precision, and accurately annotated 3D facial database using a structured light system. Building upon this foundation, we proposed a novel point cloud sampling method and 3D facial landmark detection algorithm. This method utilizes a curvature-fused graph attention network to directly predict landmark coordinates from 3D point clouds. Firstly, we extracted a simplifying point set carrying curvature information from the original 3D facial point cloud via geometric point sampling. Then, we incorporated curvature-encoded positional information as the learning component of the attention module and employed it as a feature extractor to construct the network. We evaluated the performance of the method on BU-3DFE, FaceScape and our custom dataset. Compared to existing facial landmark detection algorithms, our model achieved higher accuracy. To facilitate future research on face related applications, we have made the database available on Github 1.
- 26 July 2025
- Published by Elsevier in ISPRS Journal of Photogrammetry and Remote Sensing
- Vol. 228, 453-466
High-precision underwater turbidity removal for 3D structured light reconstruction using division-of-focal-plane polarization
- 26 July 2025
- Published by Elsevier in ISPRS Journal of Photogrammetry and Remote Sensing
- Vol. 228, 453-466
High-precision underwater turbidity removal for 3D structured light reconstruction using division-of-focal-plane polarization
- Open Access
- 7 July 2023
- Published by Association for Computing Machinery (ACM)
- p. 19-25
A Food Package Recognition and Sorting System Based on Structured Light and Deep Learning
Vision algorithm-based robotic arm grasping system is one of the robotic arm systems that can be applied to a wide range of scenarios. It uses algorithms to automatically identify the location of the target and guide the robotic arm to grasp it, which has more flexible features than the teachable robotic arm grasping system. However, for some food packages, their transparent packages or reflective materials bring challenges to the recognition of vision algorithms, and traditional vision algorithms cannot achieve high accuracy for these packages. In addition, in the process of robotic arm grasping, the positioning on the z-axis height still requires manual setting of parameters, which may cause errors. Based on the above two problems, we designed a sorting system for food packaging using deep learning algorithms and structured light 3D reconstruction technology. Using a pre-trained MASK R-CNN model to recognize the class of the object in the image and get its 2D coordinates, then using structured light 3D reconstruction technique to calculate its 3D coordinates, and finally after the coordinate system conversion to guide the robotic arm for grasping. After testing, it is shown that the method can fully automate the recognition and grasping of different kinds of food packages with high accuracy. Using this method, it can help food manufacturers to reduce production costs and improve production efficiency.
- 7 July 2023
- Published by Association for Computing Machinery (ACM)
- p. 19-25
Open AccessA Food Package Recognition and Sorting System Based on Structured Light and Deep Learning
Vision algorithm-based robotic arm grasping system is one of the robotic arm systems that can be applied to a wide range of scenarios. It uses algorithms to automatically identify the location of the target and guide the robotic arm to grasp it, which has more flexible features than the teachable robotic arm grasping system. However, for some food packages, their transparent packages or reflective materials bring challenges to the recognition of vision algorithms, and traditional vision algorithms cannot achieve high accuracy for these packages. In addition, in the process of robotic arm grasping, the positioning on the z-axis height still requires manual setting of parameters, which may cause errors. Based on the above two problems, we designed a sorting system for food packaging using deep learning algorithms and structured light 3D reconstruction technology. Using a pre-trained MASK R-CNN model to recognize the class of the object in the image and get its 2D coordinates, then using structured light 3D reconstruction technique to calculate its 3D coordinates, and finally after the coordinate system conversion to guide the robotic arm for grasping. After testing, it is shown that the method can fully automate the recognition and grasping of different kinds of food packages with high accuracy. Using this method, it can help food manufacturers to reduce production costs and improve production efficiency.
- Open Access
- 26 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 983-989
A Few-shot Learning Method for the Defect Inspection of Lithium Battery Sealing Nails
Vision-based industrial surface defect detection utilizing computer vision technologies to analyze defects in the appearance of industrial products has become popular in intelligent manufacturing. It makes inspectors move away from inefficient and labor-consuming traditional inspection methods. In this field, sealing nails play a vital role in the power battery of vehicles, and the industrial piece needs strict quality inspection according to its visual appearance before application. However, many difficulties exist, such as the lack of defect samples, low visibility of defects, and irregular shapes in the defect detection of industrial sealing nails. In this paper, we first re-labeled all non-normal areas based on the geometric contour features of the defects and made a practical classification. Second, obtain multi-dimensional image information by the polarization imaging technique; thus, it can effectively cope with low visibility. Third, proposing a new context-based Copy-Paste augmentation approach that can effectively expand the sealing nail dataset and improve the segmentation accuracy. Several experimental results have proven our methods’ accuracy and feasibility in segmentation detection models. For example, the mean pixel accuracy(mPA) criteria enhanced by about 14.9% compared with traditional methods.
- 26 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 983-989
Open AccessA Few-shot Learning Method for the Defect Inspection of Lithium Battery Sealing Nails
Vision-based industrial surface defect detection utilizing computer vision technologies to analyze defects in the appearance of industrial products has become popular in intelligent manufacturing. It makes inspectors move away from inefficient and labor-consuming traditional inspection methods. In this field, sealing nails play a vital role in the power battery of vehicles, and the industrial piece needs strict quality inspection according to its visual appearance before application. However, many difficulties exist, such as the lack of defect samples, low visibility of defects, and irregular shapes in the defect detection of industrial sealing nails. In this paper, we first re-labeled all non-normal areas based on the geometric contour features of the defects and made a practical classification. Second, obtain multi-dimensional image information by the polarization imaging technique; thus, it can effectively cope with low visibility. Third, proposing a new context-based Copy-Paste augmentation approach that can effectively expand the sealing nail dataset and improve the segmentation accuracy. Several experimental results have proven our methods’ accuracy and feasibility in segmentation detection models. For example, the mean pixel accuracy(mPA) criteria enhanced by about 14.9% compared with traditional methods.