Sep. 2025: Two co-authored papers accepted by NeurIPS 2025 (one spotlight)
Aug. 2025: Paper titled 'Task-Aware Parameter-Efficient Fine-Tuning of Large Pre-Trained Models at the Edge' accepted by IEEE GLOBECOM 2025
Apr. 2025: Paper titled 'R-ACP: Real-Time Adaptive Collaborative Perception Leveraging Robust Task-Oriented Communications' accepted by IEEE JSAC
Aug. 2025: Awarded Outstanding Academic Performance Award (OAPA) Scholarship by City University of Hong Kong
Mar. 2025: Received IEEE Robotics and Automation Society (RAS) Student Grant
Sep. 2025: Passed PhD Qualifying Examination (QE)
Research Experience
Conducts research at the Wireless Intelligence and Networked Things Laboratory (WINET) and JC STEM Lab of Smart City
Works on efficient system design for connected autonomous driving, including task partitioning and communication scheduling
Investigates real-world data fabrication attacks (e.g., cross-view and adaptive perturbations) and develops defense mechanisms for BEV and multi-stage pipelines
Performs street view synthesis and data augmentation to generate rare and long-tail scenarios for improved model generalization
Builds temporally consistent, interactive world models with 3D trajectory prediction for testing and real-time decision support
Background
Research interests include collaborative perception, autonomous driving, and AI security
Dedicated to designing robust and efficient systems for next-generation connected autonomous driving
Focuses on system efficiency and security in multi-agent connected autonomous driving
Leverages generative models for data augmentation and interactive world models to enhance generalization and decision-making
Serves as a program committee member for ICML, ICLR, ACM MM, and reviewer for IEEE TITS, TMC, RA-L, etc.