About Me
I am a Research Fellow at the Auckland Bioengineering Institute, The University of Auckland, where I lead two externally funded projects as PI. I am also Principal AI Engineer at Determ Technology, where I lead the development of Vision-Language-Action (VLA) models for robotics. I received my Ph.D. in Intelligent Science and Technology from Tongji University in 2025. My research sits at the intersection of computer vision, embodied AI, causal inference, and biomedical engineering.
Research Interest
My current research focuses on 3D spatial understanding, causal inference, and embodied AI. I am especially interested in intelligent robotic systems that can perceive, reason, and act safely in complex physical and clinical environments.
Supported by an HRC Explorer Grant, I develop physics-constrained AI for marker-less tumour localisation in breast surgery. The approach embeds biomechanical models of soft-tissue deformation to track the tumour in the deformed breast with calibrated uncertainty. It builds on my work on real-time soft-tissue segmentation, reconstruction, and deformation-aware tracking for mixed-reality surgical guidance. I also design LLM-based multi-agent systems for smart glasses that combine reasoning with physiological sensing for context-aware health feedback.
In an industry collaboration with the world's second-largest short-video platform, I combine small world models with causal inference for large-scale recommender systems, estimating supply value and cannibalization effects.
At Determ Technology, I build a closed-loop engineering framework for VLA models in robotic manipulation. It covers data curation, automated data auditing and annotation, human-in-the-loop RL + VLA training, and real-robot RGB/RGB-D deployment. Part of this work is released as the open-source Loop-Engineering-for-VLA toolkit (700+ GitHub stars).
Recent Publications
- ALM-MTA: Front-Door Causal Multi-Touch Attribution Method for Creator-Ecosystem Optimization. Y. Liu, , H. Liu, J. Liang, H. Li, K. Gai, and H. Bai. International Conference on Learning Representations (ICLR), 2026.
- Direct Routing Gradient (DRGrad): A Personalized Information Surgery for Multi-Task Learning Recommendations. Y. Liu, Y. Miao, and . AAAI Conference on Artificial Intelligence, 2025.
- Augmented reality and indoor positioning based mobile production monitoring system to support workers with human-in-the-loop. , J. Lu, Y. Lu, H. Zhang, Y. Fan, and Z. Zhang. Robotics and Computer-Integrated Manufacturing, 2024.
- Semantic knowledge-driven A-GASeq: A dynamic graph learning approach for assembly sequence optimization. , J. Lu, Y. Lu, W. Gao, Y. Fan, Y. Xu, and H. Zhang. Computers in Industry, 2024.
Recent Honors and Grants
- 2026-2028, Health Research Council of New Zealand (HRC) Explorer Grant, PI, Physics-Constrained AI for Marker-less Tumour Localisation in Breast Surgery, New Zealand.
- 2026, Gold Reviewer Reward, International Conference on Machine Learning (ICML).
- 2026-2027, Research Grant from Beijing Dajia Internet Information Technology Co., Ltd., PI, Supply Value and Cannibalization Effect of Recommendation System.
- 2025, ABI research project fund (Consolidate), AI, Foundation Models for Dynamic Human-Space Understanding in Clinical Environments, New Zealand.
- 2025, ABI Rapid-test project fund, PI, Force-Augmented Visual Tracking for Soft Tissue Deformation in Mixed-Reality Surgical Guidance.
- 2024-2025, Industry-Sponsored Research Grant, PI, Development of a Large Language Model Bridge Agent, China.
- 2025, Excellent Doctoral Dissertation, Tongji University.
Experience
- Oct 2026-Present, Research Fellow, Auckland Bioengineering Institute, The University of Auckland, New Zealand.
- Jan 2026-Present, Principal AI Engineer (AI Scientist until Jul 2026), Determ Technology, Auckland, New Zealand.
- Jul 2025-Sep 2026, Postdoctoral Fellow, Auckland Bioengineering Institute, The University of Auckland, New Zealand.
- Dec 2022-Feb 2025, Visiting Researcher, The University of Auckland, New Zealand.
- Oct 2018-Sep 2019, AI Engineer (Intern), China UnionPay, Shanghai, China.
Education
- 2020-2025, Ph.D. in Intelligent Science and Technology, Tongji University, China.
- 2017-2020, M.Eng. in Control Engineering, Tongji University, China. Phoenix Scholarship (Phoenix Contact, Germany).