Jiahua Dong

About

Jiahua Dong is a leading researcher in robotic perception, specializing in 3D vision, visual-tactile fusion, and lifelong learning systems. His most impactful contribution is **LinK3D**, a linear keypoints representation method for 3D LiDAR point clouds that has garnered 59 citations since 2024, addressing critical challenges in feature extraction and matching for robotic tasks like object detection and registration. Dong has also pioneered **partial visual-tactile fused learning** for robotic object recognition (40 citations), bridging the gap between vision and touch to enhance perception in unstructured environments. His work on **lifelong robotic visual-tactile perception learning** (34 citations) advances open-ended, adaptive learning systems that can continuously acquire knowledge without catastrophic forgetting—a key challenge in autonomous robotics. Notable achievements include developing **open-ended online learning** frameworks for autonomous visual perception and **incremental 3D object detection** methods that enable robots to learn new classes over time. Dong’s research directly impacts autonomous driving, robotic manipulation, and augmented reality, making him a pivotal figure in creating robots that perceive, learn, and adapt like humans.

Research Focus

Key Achievements

Top Papers

1. 1
   LinK3D: Linear Keypoints Representation for 3D LiDAR Point Cloud

   59 citations · 2024
2. 2
   Partial Visual-Tactile Fused Learning for Robotic Object Recognition

   40 citations · 2021
3. 3
   Lifelong robotic visual-tactile perception learning

   34 citations · 2021
4. 4
   Open-Ended Online Learning for Autonomous Visual Perception

   7 citations · 2023
5. 5
   LinK3D: Linear Keypoints Representation for 3D LiDAR Point Cloud

   7 citations · 2022
6. 6
   Interactive mechanical arm control system based on Kinect

   4 citations · 2016
7. 7
   I3DOD: Towards Incremental 3D Object Detection via Prompting

   2 citations · 2023

Key Collaborators

Contact & Links