Lei Zhou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lei Zhou has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Haozhe Wang
National University of Singapore
2 shared publications - Marcelo H. Ang
Department of Mechanical Engineering, National University of Singapore, Singapore
2 shared publications - Zhiyang Liu
National University of Singapore
2 shared publications Adn Marcelo H. Ang.
The Advanced Robotics Centre National Univer- sity of Singapore 117608 Singapore
1 shared publicationBin Cao
State Key Laboratory of Robotics and Systems, School of Mechatronics Engineering, Harbin Institute of Technology (HIT), Harbin, China
1 shared publication- Chao Wang
State Key Laboratory of Robotics and Systems, School of Mechatronics Engineering, Harbin Institute of Technology (HIT), Harbin, China
1 shared publication
Latest Publications
- 29 September 2024
- Published by Springer Nature
- p. 229-234
Enhancing Pixel-Wise Robotic Grasping with Residual Blocks
- 29 September 2024
- Published by Springer Nature
- p. 229-234
Enhancing Pixel-Wise Robotic Grasping with Residual Blocks
- 21 April 2024
- Published by Springer Nature
- p. 31-44
EA-Repose: Efficient and Accurate Feature-Metric-Based 6D Object Pose Refinement via Deep Reinforcement Learning
- 21 April 2024
- Published by Springer Nature
- p. 31-44
EA-Repose: Efficient and Accurate Feature-Metric-Based 6D Object Pose Refinement via Deep Reinforcement Learning
- Open Access
You Only Scan Once: A Dynamic Scene Reconstruction Pipeline for 6-DoF Robotic Grasping of Novel Objects
In the realm of robotic grasping, achieving accurate and reliable interactions with the environment is a pivotal challenge. Traditional methods of grasp planning methods utilizing partial point clouds derived from depth image often suffer from reduced scene understanding due to occlusion, ultimately impeding their grasping accuracy. Furthermore, scene reconstruction methods have primarily relied upon static techniques, which are susceptible to environment change during manipulation process limits their efficacy in real-time grasping tasks. To address these limitations, this paper introduces a novel two-stage pipeline for dynamic scene reconstruction. In the first stage, our approach takes scene scanning as input to register each target object with mesh reconstruction and novel object pose tracking. In the second stage, pose tracking is still performed to provide object poses in real-time, enabling our approach to transform the reconstructed object point clouds back into the scene. Unlike conventional methodologies, which rely on static scene snapshots, our method continuously captures the evolving scene geometry, resulting in a comprehensive and up-to-date point cloud representation. By circumventing the constraints posed by occlusion, our method enhances the overall grasp planning process and empowers state-of-the-art 6-DoF robotic grasping algorithms to exhibit markedly improved accuracy.
Open AccessYou Only Scan Once: A Dynamic Scene Reconstruction Pipeline for 6-DoF Robotic Grasping of Novel Objects
In the realm of robotic grasping, achieving accurate and reliable interactions with the environment is a pivotal challenge. Traditional methods of grasp planning methods utilizing partial point clouds derived from depth image often suffer from reduced scene understanding due to occlusion, ultimately impeding their grasping accuracy. Furthermore, scene reconstruction methods have primarily relied upon static techniques, which are susceptible to environment change during manipulation process limits their efficacy in real-time grasping tasks. To address these limitations, this paper introduces a novel two-stage pipeline for dynamic scene reconstruction. In the first stage, our approach takes scene scanning as input to register each target object with mesh reconstruction and novel object pose tracking. In the second stage, pose tracking is still performed to provide object poses in real-time, enabling our approach to transform the reconstructed object point clouds back into the scene. Unlike conventional methodologies, which rely on static scene snapshots, our method continuously captures the evolving scene geometry, resulting in a comprehensive and up-to-date point cloud representation. By circumventing the constraints posed by occlusion, our method enhances the overall grasp planning process and empowers state-of-the-art 6-DoF robotic grasping algorithms to exhibit markedly improved accuracy.
- 5 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 23 (22) , 27624-27633
LWOSNet: A Lightweight One-Shot Network Framework for Object Pose Estimation
The 6D pose estimation of objects is a crucial task for robotic manipulation. The currently popular methods, i.e., deep learning-based methods, usually have high requirements on the training dataset and the network architecture, which is likely to increase the cost of data annotation and training time. In this paper, we propose a lightweight one-shot network (LWOSNet) to estimate the 6D poses of multiple objects in real time and provide two feasible routes to generate synthetic training data with the objects at hand. The input of LWOSNet is a red-green-blue (RGB) image, and the output is the objects’ semantic labels and 6D poses. The whole process is divided into three stages: the image pre-processing stage, the keypoints extraction stage, and the 6D pose inference stage. Firstly, we leverage the first eight layers of visual geometry group 19(VGG-19) and two convolutional layers to downscale the dimensionality of the image feature, which effectively reduces the parameters of the network. Then, the processed features are input into two different network branches to identify the categories of the objects and generate the 3D bounding boxes. Finally, the LWOSNet outputs the semantic labels and the 6D poses calculated by the perspective-n-point (PnP) algorithm. Additionally, we conducted a series of detection experiments and robot grasping experiments. The experimental results indicate that the LWOSNet accurately detects the categories and 6D poses of multiple objects, and the robot successfully grasps the target objects based on this information.
- 5 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 23 (22) , 27624-27633
LWOSNet: A Lightweight One-Shot Network Framework for Object Pose Estimation
The 6D pose estimation of objects is a crucial task for robotic manipulation. The currently popular methods, i.e., deep learning-based methods, usually have high requirements on the training dataset and the network architecture, which is likely to increase the cost of data annotation and training time. In this paper, we propose a lightweight one-shot network (LWOSNet) to estimate the 6D poses of multiple objects in real time and provide two feasible routes to generate synthetic training data with the objects at hand. The input of LWOSNet is a red-green-blue (RGB) image, and the output is the objects’ semantic labels and 6D poses. The whole process is divided into three stages: the image pre-processing stage, the keypoints extraction stage, and the 6D pose inference stage. Firstly, we leverage the first eight layers of visual geometry group 19(VGG-19) and two convolutional layers to downscale the dimensionality of the image feature, which effectively reduces the parameters of the network. Then, the processed features are input into two different network branches to identify the categories of the objects and generate the 3D bounding boxes. Finally, the LWOSNet outputs the semantic labels and the 6D poses calculated by the perspective-n-point (PnP) algorithm. Additionally, we conducted a series of detection experiments and robot grasping experiments. The experimental results indicate that the LWOSNet accurately detects the categories and 6D poses of multiple objects, and the robot successfully grasps the target objects based on this information.