A5004211515- Robot Manipulation and Learning
- Human Pose and Action Recognition
- Robotics and Sensor-Based Localization
- 3D Surveying and Cultural Heritage
- Infrared Target Detection Methodologies
- Impact of Light on Environment and Health
- Remote Sensing in Agriculture
- Advanced Image Fusion Techniques
- 3D Shape Modeling and Analysis
- Video Surveillance and Tracking Methods
- Reinforcement Learning in Robotics
- Optical measurement and interference techniques
- Advanced Vision and Imaging
- Robotic Locomotion and Control
Örebro University
2021-2024
Tongji University
2024
Modeling dynamics of deformable linear objects (DLOs), such as cables, hoses, sutures, and catheters, is an important challenging problem for many robotic manipulation applications. In this paper, we propose the first method to model learn full 3D DLOs from data. Our approach capable capturing complex twisting bending allows local effects propagate globally. To end, adapt interaction network (IN) learning between neighboring segments in a DLO augment it with recurrent propagating along...
Robots manipulating deformable linear objects (DLOs) – such as surgical sutures in medical robotics, or cables and hoses industrial assembly can benefit substantially from accurate fast differentiable predictive models. However, the off-the-shelf analytic physics models fall short of differentiability. Recently, neural-network-based data-driven have shown promising results learning DLO dynamics. These additional advantages compared to models, they are be used gradient-based trajectory...
Tracking of deformable linear objects (DLOs) is important for many robotic applications. However, achieving robust and accurate tracking challenging due to the lack distinctive features or appearance on DLO, object's high-dimensional state space, presence occlusion. In this letter, we propose a method DLO by applying particle filter approach within lower-dimensional embedding learned an autoencoder. The dimensionality reduction preserves variation, while simultaneously enabling accurately...
Traditional approaches to manipulating the state of deformable linear objects (DLOs) - i.e., cables, ropes rely on model-based planning. However, constructing an accurate dynamic model a DLO is challenging due complexity interactions and high number degrees freedom. This renders task achieving desired shape particularly difficult motivates use model-free alternatives, which while maintaining generality suffer from sample complexity. In this paper, we bridge gap between these fundamentally...
Branched deformable linear objects (BDLOs), such as wire harnesses, are important connecting components in manufacturing industry. However, due to deformability, the lack of distinct visual features, and complex branched structure, automating tasks involving these BDLOs remains a challenge. In this paper, we propose particle-filter-based method track state BDLO. To circumvent high cost tracking high-dimensional BDLO state, instead each branch an individual B-spline curve. addition,...