Xingshuo Jing
Papers
10
Total Citations
141
H-Index
7
About
Xingshuo Jing is a leading researcher in robotic manipulation, with a focus on bridging the sim-to-real gap for dexterous grasping and tactile sensing. Their work centers on three key areas: grasp pose detection under task constraints, skill generalization through tactile learning, and domain adaptation for visual and tactile sensors. Jing's major contributions include developing affordance-based task constraint learning for single-view point cloud grasping (33 citations) and pioneering unsupervised adversarial domain adaptation for transferring optical tactile skills from simulation to reality (22 citations). Their research on skill generalization of tubular object manipulation using tactile sensing and Sim2Real learning has garnered 30 citations, demonstrating significant impact in enabling robots to handle complex, deformable objects. Notably, Jing has advanced learning from demonstration techniques, including hierarchical pick-and-place manipulation from under-specified human demonstrations (10 citations) and video-to-command translation using two-stream 2-D/3-D residual networks with self-attention (8 citations). Their recent work on pixel-level domain adaptation for real-to-sim object pose estimation (7 citations) and depth completion for transparent objects (5 citations) continues to push boundaries in reducing domain gaps. With over 140 total citations across ten publications, Jing's research is instrumental in making robotic manipulation more adaptable, efficient, and deployable in real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7Pixel-Level Domain Adaptation for Real-to-Sim Object Pose Estimation7 citations · 2023
- 8
- 9
- 10