Hi, I'm Longxiaoze, a computer vision student. And I have a lovely girlfriend.

🔬 I am currently at the CNRS-AIST JRL Humanoid Lab at the University of Tsukuba, focusing on visual SLAM research.

📚 I’m currently learning 3DGS, V-slam and Unitree humanoid robots.

💬 Ask me about Here

• PLP-SLAM(Doing) and using Point clouds as prior maps to relocalization(2D-3D-pose-tracking-update Done)
• Update PLP-SLAM to stella_vslamstella_plp_vslam
• 360° panoramic camera projects insta360X4 publish humble and Insta360 ROS driver for insta360 X4

• Object Detection on KITTI Autonomous Driving Scenes Based on YOLOv5 and PointNet(Junior year)
• SLAM14 NOTE
• slam_in_autonomous_driving_notes.
• Using Point clouds as prior maps to relocalization(2D-3D-pose-tracking-update)
• Update DSL(Monocular Direct Sparse Localization in a Prior 3D Surfel Map) to ubuntu20+cuda11.8+opencv4. DSL2
• slam_in_autonomous_driving_notes
• 3DGS-based SLAM
• VSLAM+DLfeaturePoint
• Vggt-based project (CVPR2025 best paper) (Align the Vggt results from camera scale to the world scale.)
• Fast-livo2 Ros2 humble and handhold supported for Ros2
• Unitree G1 learning shenlan and note
• 360°-panoramic-camera-based Gaussian splatting-based dense reconstruction. PGSR

• Update PLP-SLAM to stella_vslam
• Stella vslam
• 360 projects
• slam_in_autonomous_driving book(gaoxiang) and write my slam_in_autonomous_driving_notes

• TechShare Co., Ltd.（2025.10.2 – now） — Robots Perception (Mobile robots + Humanoid robots)

  • Development of mobile robots based on 360° panoramic cameras
  • LVI-SLAM (LiDAR-Visual-Inertial SLAM): Development of the Fast-LIVO2 ROS2 Humble version

• DJI JAPAN Co., Ltd.（2025.9.2 – 2025.9.31） — Camera System Development and Optimization

  • Optimized camera system compilation speed (by over 40%)
  • Developed functionality for specific scripts

• Turing Inc.（2025.4.2 – 2025.6.27） — Autonomous Driving

  • VGGT-based real-world-scale monocular camera estimation and simulated LiDAR point-cloud generation
  • LiDAR and monocular 3D object detection + localization techniques based on Visual SLAM / COLMAP

• VRC Inc.（2022.8.3 – 2023.4.3） — Human 3D Reconstruction

  • Operation of human 3D avatar algorithms
  • Optimization of 3D reconstruction algorithms

• DeepGlint Information Technologies Co., Ltd.（2021.11.15 – 2022.4.15） — AI Computer Vision

  • Object detection (mmDetection), image unsupervised anomaly detection
  • Assisted in developing intelligent cars (ROS/V-SLAM) and robot arms (MoveIt)

• image object detection
• 3D mono object detection
• ros+moveit(just use)
• V-slam (orbslam2+orbslam3)
• image anomaly detection
• Learn the basics of NERF and 3DGS.
• Learn the basics of C++.
• Japanese N2 passed(2024.07)
• Japanese Driver's License(2024.11)

⭐️ From Longxiaoze