Pinlong Cai | 蔡品隆

Research Scientist, Shanghai Artificial Intelligence Laboratory.
701 Yunjin Road in Xuhui District, Shanghai, China
caipinlong@pjlab.org.cn
Google scholar | ORCID | ResearchGate
I received my Ph.D. degree in Traffic Information Engineering and Control from Beihang University, Beijing, China, supervised by Prof. Yunpeng Wang and Prof. Guangquan Lu. From 2016 to 2017, I was a Research & Development Engineer with the Quanzhou Institute of Equipment Manufacturing, Haixi Institutes, Chinese Academy of Sciences, led by Prof. Hao Chen. From 2020 to 2021, I was a Standard & Strategy Engineer at ZTE Corporation. Currently, I am a Research Scientist at Shanghai Artificial Intelligence Laboratory.
I have already published more than 40 academic papers in peer-reviewed journals and conferences (including 1 ESI Top 1% highly cited paper), obtained more than 10 patents, and 1 software copyright. I am also the reviewer of more than 50 journals and conferences. I participate in writing books of Urban Transport System Operation Reliability Analysis Method and Intelligent Road Transport Systems: An Introduction to Key Technologies. I also participated in the preparation of the Data Compliance Guidelines for Intelligent Connected Vehicles, which was released at the 2022 World Artificial Intelligence Conference (WAIC). I am the winner of the Yangfan Special Project of the Shanghai Qimingxing Program.
My research interests include multimodal large model, knowledge-driven inference, and cooperative vehicle infrastructure system.
news
Sep 3, 2025 | HetaRAG, a hybrid, deep-retrieval RAG framework that unifies multiple heterogeneous data stores, has been released (Code) |
---|---|
Aug 14, 2025 | One paper titled “LeanRAG: Knowledge-Graph-Based Generation with Semantic Aggregation and Hierarchical Retrieval” has been summited to the arXiv.org e-Print archive |
Jun 1, 2025 | One paper titled “KG-TRACES: Enhancing Large Language Models with Knowledge Graph-constrained Trajectory Reasoning and Attribution Supervision” has been summited to the arXiv.org e-Print archive |
Sep 27, 2024 | LimSimLight (new version of LimSim to parsing OpenDrive map files and simulate by the new self-developed engine) has been released (Code) |
Apr 29, 2024 | InternVL 1.5 (an open-source multimodal large language model) has been released (Code, Paper, Model and Dataset) |
Mar 31, 2024 | LimSim++: A Closed-Loop Platform for Deploying Multimodal LLMs in Autonomous Driving was accepted by IEEE IV 2024 (Code and Paper) |
Dec 24, 2023 | One paper titled “How drivers perform under different scenarios: Ability-related driving style extraction for large-scale dataset” has been accepted for publication in the Accident Analysis & Prevention |
Dec 12, 2023 | One paper titled “Towards Knowledge-driven Autonomous Driving” has been summited to the arXiv.org e-Print archive |
Jul 15, 2023 | A Long-term Interactive Multi-scenario traffic Simulator (LimSim) is released (Code and Paper) |
Sep 2, 2022 | The release of the Guidelines on Data Compliance of Connected Automated Vehicles in WAIC 2022 |
selected publications
-
TITSIEEE Transactions on Intelligent Transportation SystemsIn IEEE Transactions on Intelligent Transportation Systems, 2025
-
WACVDrive like a human: Rethinking autonomous driving with large language modelsIn Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024
-
AAPHow drivers perform under different scenarios: ability-related driving style extraction for large-scale datasetAccident Analysis & Prevention, 2024
-
JTEATrajectory guidance for connected human-driving vehicles through the interactions between drivers and roadside unitsJournal of transportation engineering, Part A: Systems, 2023
-
ITSMHuman-Like Decision Making at Unsignalized Intersections Using Social Value OrientationIEEE Intelligent Transportation Systems Magazine, 2024
-
ITSCLimSim: A long-term interactive multi-scenario traffic simulatorIn 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), 2023