cv

Bachelor's Degree in Information Engineering (Electronic and Electrical Engineering)

• GPA: 3.91/4.0
• Ranking: 2/46
• Enrolled in Advanced Honors Program of Dept. of EEE. (Wo'loong Class)
• Major Coursework
  • Robotics Motion and Control
  • Robotics Perception and Intelligence
  • Digital Signal Processing
  • Signals and Systems
  • Analog Circuits
  • Digital Circuits
  • Engineering Mathematics
  • Probability and Statistics
  • General Physics
  • Calculus
  • Linear Algebra

AREA3D: Active Reconstruction Agent with Unified Feed-Forward 3D Perception and Vision-Language Guidance

• Authors: Tianling Xu, Shengzhe Gan, Leslie Gu, Yuelei Li, Fangneng Zhan, Hanspeter Pfister.

Remote Research Intern

• Advisor: Dr. Fangneng Zhan
• AREA3D Framework Development: Developed an active reconstruction agent that unifies feed-forward geometric confidence with VLM-based semantic guidance to resolve occlusions and ambiguities.
• SOTA Baseline Benchmarking: Reproduced and optimized complex baselines including FisherRF and Air-Embodied.
• 3DGS Experimental Validation: Conducted large-scale 3D Gaussian Splatting experiments on Replica and OmniObject3D datasets, achieving SOTA PSNR/SSIM.
• Active Perception Strategy: Implemented 'Dual-Field' uncertainty fusion using Monte Carlo ray sampling and frustum-based decay.

On-Site Research Assistant

• Advisor: Prof. Mengxue Hou
• Robotics System Engineering: Architected a complete grasping system (YOLO, SAM, GraspNet) within a ROS2 framework to achieve autonomous manipulation on a Kinova Gen3 arm.
• Novel 3D Representation for Grasping: Enhanced 3D Gaussian Splatting by embedding a 'graspness' metric into primitives to guide next-best-view selection.
• Advanced Perception Research: Prototyped models with PointNet++ and Point Transformers to learn continuous Signed Distance Fields (SDFs) from partial point clouds.

Undergraduate Researcher

• Topic: Reflection Awareness in SLAM system
• Advisor: Prof. Hong Zhang
• Multi-Modal SLAM Fusion: Augmented Fast-LIO with a vision-based module to detect mirrors and filter reflective LiDAR points via 3D-to-2D projection.
• Real-Time Point Cloud Filtering: Developing a real-time ROS node to filter reflective artifacts from LiDAR scans.
• Quantitative Validation: Benchmarked the enhanced SLAM system against baselines, validating improved trajectory accuracy and map quality.

Chinese Chess HCI Robot

• Full System Integration: Engineered an end-to-end chess robot in ROS with KINOVA Gen3 Lite arm and RealSense camera.
• Computer Vision & AI: Deployed custom-trained YOLOv8s for real-time piece recognition.
• Robotic Motion Planning: Implemented custom optimization-based IK solver and cubic polynomial trajectory planner for positional errors under 10^-5m.

Autonomatic SLAM Mobile Robot

• Wireless Communication: Developed wireless TCP/IP bridge using ESP32 to stream sensor data to ROS.
• LiDAR & Odometry Fusion SLAM: Deployed Hector SLAM for real-time 2D mapping fusing LiDAR and odometry.
• ROS Data Pipeline: Engineered data pipeline to parse sensor streams into standard topics (/scan, /tf, /odom).

• First Class of the Merit Student Scholarship (Academic Year 2024-2025)
• Honorary Title of Outstanding Student
• Mathematical Competition in Modeling (MCM) - Meritorious Winner

• First Class of the Merit Student Scholarship (Academic Year 2023-2024)
• Honorary Title of Outstanding Student
• First Prize of the China Undergraduate Physics Tournament (CUPT)
• Enrolled by Advanced Honor Program (Top 9 students in Dept. EEE)