Zirui Li, Ph.D.

AI4X Postdoc Fellow at Nanyang Technological University

About Me

Zirui Li, Ph.D.

I am a AI4X Postdoctoral Researcher in AutoMan Lab at NTU, which is led by Prof. Chen Lv and Prof. Ziwei Liu. I obtained my Ph.D. at Beijing Institute of Technology, Beijing, China, under the supervision of Prof. Jianwei Gong. Prior to this, I was a visiting researcher in TU Delft and TU Dresden from June, 2021 to June, 2024, working with Prof. Meng Wang, Prof. Victor Knoop. I am an imcoming Marie Skłodowska-Curie Actions Postdoctoral Fellowships at EPFL, colloborating with Prof. Andrea Cavallaro.

     

    Research Interests

    My research interests include:

    • Data-driven interactive behaviour modelling and motion planning in dense environments
    • Continual/lifelong learning in robotics (e.g., automated vehicles)
    • Traffic safety assessment
    • Collective intelligence emergence in edge (Postdoctoral research)

    News

    • June. 2026: My dissertation has received the IEEE ITS Best Dissertation Award-The 1st Place from the IEEE ITSS.
    • May. 2026: One paper about continual learning stability-plasticity dilemma is accepted by T-NNLS, check paper and code here.
    • Apr. 2026: One paper about continual learning in automated vehicles is accepted by IEEE T-ITS, check paper and code here.
    • March. 2026: One paper about early-stage takeover in shared-control driving is accepted by CHI 26, check check paper here.
    • Feb. 2026: I am awarded Marie Skłodowska-Curie Actions Postdoctoral Fellowships, host by EPFL.
    • Jan. 2026: I am awarded the prestigious NTU-NRF AI4X fellowship, equipped with 250K SGD fund.
    • Sep. 2025: My doctoral thesis is selected as the best Ph.D Thesis by SAE China.
    • Sep. 2025: I join AutoMan Lab at NTU as Postdoc Research Fellow.
    • Jul. 2025: I was recognised as Excellent Oral Presentation in the CASE Doctoral Student forum.
    • Jun. 2025: One paper accepted by ITSC 2025, see you all in Golden Coast.
    • May. 2025: I just successfully defensed my Ph.D. thesis.
    • May. 2025: I was recognised as the IEEE ITSS Young Professionals Fellowship.
    • Jan. 2025: I was selected for China Association for Science and Technology Youth Talent Support.
    • Apr. 2025: Welcome to my homepage.

    Biography

    Education

    Ph.D., Mechanical Engineering

    Sep. 2019 - Jun. 2025

    Beijing Institute of Technology (BIT), Beijing, China

    Advisor: Prof. Jianwei Gong

    Dissertation: Interactive Motion Planning of Automated Vehicles based on Spatiotemporal Risk Continual Learning

    Awards: Excellent Doctoral Dissertation Seedling Fund

    CSC Visiting Student

    Jun. 2022 - Jun. 2023

    TU Dresden, Germany

    Advisor: Prof. Meng Wang

    Topics: Interactive Behaviour modelling

    CSC Visiting Student

    Jun. 2021 - Jun. 2022

    TU Delft, The Netherlands

    Advisor: Prof. Meng Wang and Dr. Victor Knoop

    Topics: Interactive Behaviour modelling

    B.Eng., Xuteli Talent School (Mechanical Engineering)

    Sep. 2015 - Jun. 2019

    Beijing Institute of Technology (BIT), Beijing, China

    Advisor: Prof. Jianwei Gong

    Thesis: Analysis of influencing factors for driver decision-making at urban intersections

    Awards: Outstanding Graduate and Outstanding Undergraduate Thesis

    Work Experience

    AI4X Research Fellow

    March. 2026 - Now

    Nanyang Technological University, Singapore

    Advisor: Prof. Chen Lyu and Prof. Ziwei Liu

    Topic: CLEAR: Continual Learning for Embodied Agents via Spatio-temporal Knowledge Sharing

    Postdoc Research Fellow

    Sep. 2025 - March. 2026

    Nanyang Technological University, Singapore

    Advisor: Prof. Chen Lyu

    Topic: Collective AI

    Part-time Researcher

    Jun. 2023 - Jun. 2024

    TU Dresden, Dresden, Germany

    Advisor: Prof. Meng Wang

    Topic: DNN-based Bus/tram arrival time prediction

    Selected Research

    Below are four selected research projects I have done.

    Interactive Motion Planning for Automated Vehicles

    With the continuous advancements in information, sensing, and artificial intelligence technologies, unmanned vehicles as a new type of transport platform have made significant progress in fields such as autonomous delivery. This work focuses on modeling the interactive behaviors among multiple traffic participants, quantifying and integrating various types of risk, and designing motion planning algorithms. Accordingly, it proposes a spatiotemporal risk continual learning–based interactive motion planning method for unmanned vehicles. The “behavior prediction–risk assessment–motion planning” integrated approach aims to enhance driving safety in dynamically interactive environments through accurate spatiotemporal risk assessment.

    Related Links:

    IEEE T-ITS 1 IEEE T-ITS 2 IEEE T-ITS 3 IEEE/ASME T-Mech GEIT Nat. Sci. Data TR-Part C Demo


    Large-scale Driver Perceived Risk analysis with 140K+ ratings

    Driver's perceived risk in self-driving cars can create the very danger that automation is meant to prevent: a frightened rider may hesitate when seconds matter, misjudge hazards, or disengage. However, how this risk evolves dynamically in real world remains poorly understood. Understanding what sparks that fear is therefore critical to real-world safety. We conducted the largest controlled investigation of passenger emotion in automated driving to date. 2,164 volunteers viewed short, high-fidelity videos of four common highway situations and provided 141,628 second-by-second safety ratings.

    Related Links:

    Questionnaire Project Demo Results Demo


    Public Transport Arrival Time Prediction

    Accurate arrival time prediction (ATP) of buses and trams plays a crucial role in public transport operations. Current methods only focused on modeling the time series along the temporal dimension by investigating the point-to-point relationship between different timesteps. This strategy overlooked the latent periodic information within time series. Moreover, most studies developed algorithms for ATP based on a single or a few routes of public transport, which reduces their applicability in public transport management systems. To this end, this work proposes ArrivalNet, a two-dimensional temporal variation-based multi-step ATP for buses and trams. 125 days of public transport data from Dresden were collected for model training and validation.

    Related Links:

    IEEE T-ITS Data Sci. in Trans. Traffic Monitor Platform

    A naturalistic trajectory dataset with dense interaction

    This work presents InterHub, a curated dataset of dense multi-agent interaction events, derived from large-scale naturalistic driving recordings. The dataset is accompanied by an open-source toolkit that enables users to expand InterHub by mining additional interaction events from both public and private driving data. By offering a unified taxonomy, rich annotations, and extensible tools, InterHub supports diverse research needs-from interaction behavior modeling to algorithm benchmarking-and promotes reproducibility, scalability, and cross-dataset comparison in autonomous driving studies.

    Related Links:

    Nat. Sci. Data Video Code

    Selected Publications

     

    Preprints

    P5. Perceived risk evolution in automated driving inferred from large-scale discrete ratings
    Under Review
    X. He, Z. Li, X. Wang, R. Happee, M. Wang
    IEEE PDF
    P4. ArrivalNet: Predicting City-wide Bus/Tram Arrival Time with Two-dimensional Temporal Variation Modeling
    Under Review at OJITS
    Z. Li, P. Wolf, M. Wang
    IEEE PDF
    P3. Exploring Over-stationarization in Deep Learning-based Bus/Tram Arrival Time Prediction: Analysis and Non-stationary Effect Recovery
    Under Review at AI for Trans, 2026.
    Z. Li, B. Yang, M. Wang
    ArXiv
    P2. Complementary Learning System Empowers Online Continual Learning of Vehicle Motion Forecasting in Smart Cities
    Under Review, 2026.
    Z. Li, Y. Lin, G. Du, X. Zhao, C. Gong, C. Lu, and J. Gong
    ArXiv

     

    Selected Journal Papers

    J14. Escaping Stability-Plasticity Dilemma in Online Continual Learning for Motion Forecasting via Synergetic Memory Rehearsal
    IEEE Transactions on Neural Networks and Learning Systems, 2026.
    Y. Lin, C. Lu, T. Wu, X. Zhao, G. Du, Y. Sun, Z. Li*, J. Gong
    (* Corresponding Author)
    ArXiv
    J13. H2C: Hippocampal Circuit-inspired Continual Learning for Lifelong Trajectory Prediction in Autonomous Driving
    IEEE Transactions on Intelligent Transportation Systems, 2026.
    Y. Lin(#), Z. Li(#), G. Du, X. Zhao, C. Gong, X. Wang, C. Lu, and J. Gong
    (# Equal Contribution)
    ArXiv
    J12. Experience-Shared Variable-Step Predictive Control of Range-Extended Electric Vehicles Using Transferable Driver Model
    IEEE Transactions on Intelligent Transportation Systems, 2025, 26(1): 1123-1133.
    J Li, Z. Li*, C. Wen, Y. Wu, R. Dixon, X. Hu, H. Xu
    (* Corresponding Author)
    IEEE PDF
    J11. An Imputation-enhanced Hybrid Deep Learning Approach for Traffic Volume Prediction in Urban Networks: A Case Study in Dresden
    Data Science for Transportation, 2024, 6(3): 22.
    P Yan(#), Z. Li(#), J. Ijaradar, S. Pape, M. Körner, M. Wang
    (# Equal Contribution)
    Springer PDF
    J10. Interactive Trajectory Primitives Representation and Extraction Based on Naturalistic Driving Data (In Chinese)
    Automotive Engineering, 46.8 (2024): 1382-1393.
    Z. Li, H. Wang, J Gong, C. Lu, X. Zhao, M. Wang
    Online Website PDF
    J9. Interactive Behavior Modeling for Vulnerable Road Users With Risk-Taking Styles in Urban Scenarios: A Heterogeneous Graph Learning Approach
    IEEE Transactions on Intelligent Transportation Systems, 2024, 25(8): 8538-8555.
    Z. Li, J. Gong, Z. Zhang, C. Lu, VL. Knoop, M. Wang
    IEEE PDF
    J8. Continual Driver Behaviour Learning for Connected Vehicles and Intelligent Transportation Systems: Framework, Survey and Challenges
    Green Energy and Intelligent Transportation, 2023, 2(4): 100103.
    Z. Li, C. Gong, Y. Lin, G. Li, X. Wang, C. Lu, M. Wang, S. Chen, J. Gong
    Elsevier PDF
    J7. Leveraging Multi-Stream Information Fusion for Trajectory Prediction in Low-Illumination Scenarios: A Multi-Channel Graph Convolutional Approach
    IEEE Transactions on Intelligent Transportation Systems, 2023, 25(5): 3854-3869.
    H. Gong(#), Z. Li(#), C. Lu, G. Du, J. Gong
    (# Equal Contribution)
    IEEE Code PDF
    J6. Continual Interactive Behavior Learning with Traffic Divergence Measurement: A Dynamic Gradient Scenario Memory Approach
    IEEE Transactions on Intelligent Transportation Systems, 2023, 25(3): 2355-2372.
    Y. Lin(#), Z. Li(#), C. Gong, C. Lu, X. Wang, J. Gong
    (# Equal Contribution)
    IEEE Code PDF
    J5. Personalized driver braking behavior modeling in the car-following scenario: An importance-weight-based transfer learning approach
    IEEE Transactions on Industrial Electronics, 2022, 69(10): 10704-10714.
    Z. Li, J. Gong, C. Lu, J. Li
    IEEE PDF
    J4. A Transferable Energy Management Strategy for Hybrid Electric Vehicles via Dueling Deep Deterministic Policy Gradient
    Green Energy and Intelligent Transportation, 2022, 1(2): 100018.
    J Xu(#), Z. Li(#), G. Du, Q. Liu, L. Gao, Y. Zhao
    (# Equal Contribution)
    Elsevier Code PDF
    J3. A Hierarchical Framework for Interactive Behaviour Prediction of Heterogeneous Traffic Participants based on Graph Neural Network
    IEEE Transactions on Intelligent Transportation Systems, 2021, 23(7): 9102-9114.
    Z. Li, C. Lu, Y. Yi, J. Gong
    IEEE PDF
    J2. Interactive Behavior Prediction for Heterogeneous Traffic Participants in the Urban Road: A Graph-Neural-Network-based Multitask Learning Framework
    IEEE/ASME Transactions on Mechatronics 26.3 (2021): 1339-1349.
    Z. Li, J. Gong, C. Lu, Y. Yi
    IEEE PDF
    J1. Importance Weighted Gaussian Process Regression for Transferable Driver Behaviour Learning in the Lane Change Scenario
    IEEE Transactions on Vehicular Technology, 2020, 69(11): 12497-12509.
    Z. Li, J. Gong, C. Lu, J. Xi
    IEEE PDF

     

    Conference Papers

    C3. An Ensemble Learning Framework for Vehicle Trajectory Prediction in Interactive Scenarios
    2022 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2022: 51-57.
    Z. Li, Y. Lin, C. Gong, X. Wang, Q. Liu, J. Gong, C. Lu
    IEEE Code PDF
    C2. Transferable Driver Behavior Learning via Distribution Adaption in the Lane Change Scenario
    2019 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2019: 193-200.
    Z. Li, C. Gong, C. Lu, J. Gong, J. Lu, Y. Xu, F. Hu
    IEEE PDF
    C1. Development and Evaluation of Two Learning-based Personalized Driver Models for Pure Pursuit Path-tracking Behaviors
    2018 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2018: 79-84.
    Z. Li, B. Wang, J. Gong, T. Gao, C. Lu, G. Wang
    IEEE PDF

     

    Books

    B2. Scene Understanding and Behavior Prediction for Intelligent Vehicles (In Chinese)
    Beijing Institute of Technology Press, 2025.
    C. Lu, Z. Li, J. Gong, Y. Lin
    Purchase link
    B1. Interactive Behavior Prediction and Decision-making for Intelligent Vehicles (In Chinese)
    Beijing Institute of Technology Press, 2021.
    C. Lu, J. Gong, Z. Li
    Purchase link

    Professional Services

    Program Committee

    • Chair, Foundation Model-Empowered Decision-making and Planning for Intelligent Mobility Systems at IEEE ITSC 2026
    • Co-chair,Next-Generation Autonomous Driving Systems: Self-Evolving Decision-Making, End-to-End Architectures, and Data-Driven Evaluation at IEEE ITSC 2026
    • Co-chair, AI-Empowered Human-Machine Interaction for Intelligent Vehicles at IEEE ITSC 2026
    • Associate Editor, CIS-RAM 2026
    • Associate Editor, ITSC 2026
    • Guest Editor, Actuators
    • Guest Editor, IET-ITS
    • Working group leader, intelligent vehicle risk assessment working group under IEEE Standard Association
    • Co-chair, Cooperative Decision-making for Connected and Automated Vehicles in ITS at IEEE ITSC 2023
    • Associate Editor for IEEE IV 2023
    • Chair, Social, interactive and safe behaviors for AVs: benchmarks, models and applications at IEEE IV 2023
    • Organising committee and Proceedings editor, MFTS 2022
    • Chair, Social and Interactive Behavior Modelling in ITS at IEEE ITSC 2021

    Journal Reviewer

    • IEEE Transactions on Intelligent Transportation Systems
    • IEEE Transactions on Neural Networks and Learning Systems
    • IEEE Transactions on Pattern Analysis and Machine Intelligence
    • IEEE Transactions on Vehicular Technology
    • IEEE Transaction on Transportation Electrification
    • Accident Analysis & Prevention
    • IEEE Robotics and Automation Letters
    • IET Intelligent Transport Systems
    • Frontiers of Information Technology & Electronic Engineering
    • IEEE Transactions on Intelligent Vehicles
    • IEEE/CAA Journal of Automatica Sinica
    • IEEE Internet of Things Journal
    • Transportation Research Part C: Emerging Technologies
    • Transportation Research Part F: Traffic Psychology and Behaviour

    Conference Reviewer

    • IEEE Conference on Intelligent Transportation Systems (ITSC)
    • IEEE Intelligent Vehicles Symposium (IVs)
    • Transportation Research Board (TRB) Annual Meeting
    • IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
    • IEEE International Conference on Robotics and Automation (ICRA)

    Supervision

    • NTU: Zhiqi Mao, Yujie Yan.
    • Beijing Institute of Technology: Cheng Gong, Yunlong Lin, Chenxu Wen, Haowen Wang, Hailong Gong, Lianzhen Wei, Yangtian Yi, Xianqi He, Jingyi Xu, Yingqi Tan, Tairan Chen.
    • TU Dresden: Fanyi Wei.