I am Jianming Xing (邢鉴明), currently pursuing a second bachelor’s degree in Computer Science and Technology at Harbin Institute of Technology, Shenzhen. I received my B.Eng. in Mechanical Design, Manufacturing and Automation from Harbin Institute of Technology, Weihai.
My interests center on embodied AI, vision-language-action models, robot learning, and real-robot systems. I work across trajectory modeling, embodied spatial reasoning, robot data and inference pipelines, ROS 2 integration, and robotics software that connects learning-based policies to physical platforms.
English CV / 中文简历 / GitHub / Email
News
- Work news Joined Cytoderm as an Algorithm Engineer.
- Award news Won First Place in the CVPR 2026 ERA Workshop RoboSpatial Challenge.
- Award news Won Second Place in the CVPR 2026 ERA Workshop PointArena Challenge.
- Research news Joined iLearn at Harbin Institute of Technology, Shenzhen, as a Research Assistant.
- Education news Began a second bachelor's degree in Computer Science and Technology at Harbin Institute of Technology, Shenzhen.
Publications
Hermite-VLA Reg real-robot success demonstration
Hermite Curves as Trajectory Priors for Vision-Language-Action Models
arXiv 2026 · Under Review
Introduces piecewise cubic Hermite trajectory priors for VLA action chunks and evaluates them as discrete tokens, generative scaffolds, and training regularization across simulation and real-robot tasks.
Technical Report of RoboSpatial Challenge at CVPR 2026: Selective Reasoning Activation and Reference-Frame Disambiguation for Embodied Spatial Reasoning
CVPR 2026 ERA Workshop · RoboSpatial Challenge — 1st Place
Combines selective reasoning activation with explicit reference-frame disambiguation for embodied spatial reasoning, achieving an 80.9% overall success rate on RoboSpatial-Home.
Efficient Visual Pointing for Embodied AI: Agent-Driven Data Synthesis, Cross-Block Attention, and Iterative Correction
CVPR 2026 ERA Workshop · PointArena Challenge — 2nd Place
Builds a visual-pointing system around agent-driven data synthesis, cross-block attention, iterative coordinate correction, and category-aware expert routing, reaching 77.2% overall accuracy.
Projects
ROS2 Humble · Dobot CR5 · LeRobot v2.0 · OpenPI
VLA/OpenPI Real-Robot Data Collection and Inference Loop for CR5
Built a real-robot closed loop for Dobot CR5, Orbbec Astra2 RGB-D, and an electric gripper, covering calibration, teleoperation, synchronized data collection, dataset conversion, OpenPI fine-tuning, and WebSocket-based inference.
- Designed an HDF5 synchronization pipeline using RGB image timestamps as the master clock, aligning robot state, target action, and gripper feedback with binary search and linear interpolation.
- Reduced early camera frame drops from 27.15% to 2.57%-4.74% by using Fast DDS shared memory and RGB-only capture.
C++/Qt upper computer · U-Net seam segmentation · OpenGL digital twin
Robot Digital-Twin HMI for Weld-Seam Recognition
Built a robotics software stack for weld-seam recognition and LeArm digital-twin demos, combining semantic segmentation, Qt-based control, and real-time OpenGL visualization.
- Compared baseline U-Net, VGG-U-Net, and a VGG16-pretrained variant, improving weld-seam segmentation accuracy from 64% to 96.8% through transfer learning, data augmentation, and dataset expansion.
- Integrated image selection, Python inference calls, and segmentation-result display into a C++/Qt HMI.
Education & Experience
Education
-
–
Harbin Institute of Technology, Shenzhen
Second Bachelor's Degree in Computer Science and Technology
-
–
Harbin Institute of Technology, Weihai
B.Eng. in Mechanical Design, Manufacturing and Automation
Experience
-
–
Present
Algorithm Engineer · Cytoderm
Maintain real-robot software across Franka, Cybopal/in-house arms, and ARX platforms, including teleoperation, data collection, robot-state/action contracts, and LeRobot/RLDS conversion.
-
–
Research Assistant · iLearn, Harbin Institute of Technology, Shenzhen
Research on embodied AI, VLA trajectory priors, robot learning, and embodied spatial-reasoning challenges under the supervision of Xiang Deng.
-
–
Community Intern · Xbotics Embodied AI Community
Migrated Isaac Lab's ANYmal-C navigation task to MotrixLab NumPy/MuJoCo and rebuilt environment, terrain, reward, termination, and regression-test pipelines.
-
–
Vision Group Member · HIT Weihai HERO Robomaster Team
Contributed to Robomaster vision framework ROS 2 migration and communication development by separating image acquisition, vision processing, and communication synchronization into nodes; implemented host-MCU data reception, Topic messaging, time synchronization, and seasonal protocol and message updates.
Awards
- First Place, CVPR 2026 ERA Workshop RoboSpatial Challenge
- Second Place, CVPR 2026 ERA Workshop PointArena Challenge
- National Second Prize, 18th National Undergraduate Intelligent Vehicle Competition
- Second-class People's Scholarship, Harbin Institute of Technology (11/132)
Visitor Map
The visitor map is temporarily unavailable.
Visitor locations are approximate country or region estimates inferred by a third party from IP addresses; precise locations are not displayed.