Research Areas

At the EAR Lab, we drive the evolution of Embodied AI from principle to practice. Our research centers on creating resilient perception systems and intelligent planning algorithms that empower real-world robots. To achieve this, we actively solve foundational challenges across data-centric learning, robust multi-modal sensing, and autonomous action in complex environments.

Embodied AI simulation and real-world transfer
Robotic perception and control model development
Dataset curation for robot learning
Robotic perception and control model development

Data-Centric Embodied AI

Bridging data scarcity for real-world robots

This area addresses two core problems: i) how to train perception and action models when real robot data is scarce, and ii) how to collect and curate large-scale, robot-ready datasets.

Selected Publications

  • Yuanchuan Lai, Qing Gao, Ziyan Liang, Xianfeng Cheng, Junjie Hu, Zhaojie Ju, "DexTele: A Dual-Arm Dexterous Teleoperation System Based on Motion Retargeting and Adaptive Force Control," IEEE International Conference on Robotics and Automation (ICRA), 2026. PDF
  • Xianfeng Cheng, Qing Gao, Guangyu Chen, Rui Xiong, Junjie Hu, Yulan Guo, Zhaojie Ju, "Viper: Verifiable Imitation Learning Policy for Efficient Robotic Manipulation," IEEE International Conference on Robotics and Automation (ICRA), 2026. PDF
  • Junjie Hu, Chenyou Fan, Mete Ozay, Hua Feng, Yuan Gao, Tin Lun Lam, "Unlocking Drone Perception in Low AGL Heights: Progressive Semi-Supervised Learning for Ground-to-Aerial Perception Knowledge Transfer," IEEE Transactions on Intelligent Transportation Systems (T-ITS), 2025. PDF
  • Qiu Zheng†, Junjie Hu†, Yuming Liu, Zengfeng Zeng, Wang Fan, Tin Lun Lam, "Transferring Visual Knowledge: Semi-Supervised Instance Segmentation for Object Navigation Across Varying Height Viewpoints," IEEE International Conference on Robotics and Automation (ICRA), 2025. PDF
  • Junjie Hu, Chenyou Fan, Liguang Zhou, Qing Gao, Honghai Liu, Tin Lun Lam, "Lifelong-MonoDepth: Lifelong Learning for Multi-Domain Monocular Metric Depth Estimation," IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2025. PDF
Multi-sensor fusion depth estimation under degradation
Multi-sensor fusion depth estimation under degradation
RGB-thermal attention-enhanced fusion example
Satellite-ground fusion for 3D scene completion

Multi-Sensor Fusion

Perceiving and acting from multi-modal sources

Research on how to combine complementary sensors to improve robot perception. Sepcifically, we study how to design effective, robust, and adaptive fusion strategies for different sensor modalities, embodiments, and tasks.

Selected Publications

  • Junjie Hu, Chenyou Fan, Mete Ozay, Qing Gao, Yulan Guo, Tin Lun Lam, "Robust Depth Estimation Under Sensor Degradations: A Multi-Sensor Fusion Perspective," IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2025. PDF
  • Junjie Hu, Chenyu Bao, Mete Ozay, Chenyou Fan, Qing Gao, Honghai Liu, Tin Lun Lam, "Deep Depth Completion from Extremely Sparse Data: A Survey," IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022. PDF
  • Junjie Hu, Chenyou Fan, Xiyue Guo, Liguang Zhou, Tin Lun Lam, "Self-supervised Single-line LiDAR Depth Completion," IEEE Robotics and Automation Letters (RAL), 2023. PDF
  • Mingjian Liang†, Junjie Hu†, Chenyu Bao, Hua Feng, Fuqin Deng, Tin Lun Lam, "Explicit Attention-Enhanced Fusion for RGB-Thermal Perception Tasks," IEEE Robotics and Automation Letters (RAL), 2023. PDF
  • Xiyue Guo, Jiarui Hu, Junjie Hu, Hujun Bao, Guofeng Zhang, "SGFormer: Satellite-Ground Fusion for 3D Semantic Scene Completion," IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025. PDF
Descriptor distillation for efficient multi-robot SLAM
Semantic histogram based graph matching for multi-robot localization
Descriptor distillation for efficient multi-robot SLAM
Descriptor distillation for efficient multi-robot SLAM

Multi-Robot Collaboration

Coordinated teams for resilient robotic teamwork

Advancing collaborative robotics by enabling robots to perceive, localize, and act together in challenging environments. We mainly study multi-robot collaborative perception and planning.

Selected Publications

  • Shaobin Ling, Yun Wang, Chenyou Fan, Tin Lun Lam, Junjie Hu, "ELHPlan: Efficient Long-Horizon Task Planning for Multi-Agent Collaboration," IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2026.
  • Chenyou Fan, Junjie Hu and Jianwei Huang, "Few-Shot Multi-Agent Perception with Ranking-Based Feature Learning," IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2023. Paper
  • Xiyue Guo, Junjie Hu, Junfeng Chen, Fuqin Deng, Tin Lun Lam, "Semantic Histogram Based Graph Matching for Real-Time Multi-Robot Global Localization in Large Scale Environment," IEEE Robotics and Automation Letters (RAL), 2021. Paper
  • Xiyue Guo, Junjie Hu, Hujun Bao, Guofeng Zhang, "Descriptor Distillation for Efficient Multi-robot SLAM," IEEE International Conference on Robotics and Automation (ICRA), 2023. Paper
  • Yuan Gao, Junfeng Chen, Xi Chen, Chongyang Wang, Junjie Hu, Fuqin Deng, Tin Lun Lam, "Asymmetric Self-play Enabled Intelligent Heterogeneous Multi-robot Catching System using Deep Multi-agent Reinforcement Learning," IEEE Transactions on Robotics (TRO), 2023. Paper