PDF Version

Curriculum Vitae

Education

Publications

Industrial

Academic

Services

Skills

Education

B.Eng. in Computer Science and Technology

School of Computer Science and Engineering

Beihang University, 2019 - 2023

Publications

Industrial

University Health Network - Team Lead, Machine Learning

February 2025 - PRESENT

University Health Network (UHN) is an alliance of leading healthcare research facilities at the forefront of medical innovation, advancing cutting-edge AI research to transform healthcare and improve patient outcomes. In 2024, Toronto General Hospital, part of UHN, was ranked as the No. 3 hospital worldwide.
  • I lead the Machine Learning Research Team at the SARA Lab, UHN, overseeing research in computer vision, graphics, and reinforcement learning. I also supervise healthcare-tailored game development and web application projects.
  • Our mission is to pioneer AI-driven solutions for surgical and clinical applications.

University Health Network - Machine Learning Researcher

May 2024 - December 2024

The Surgical AI Research Academy (SARA) is dedicated towards research and innovation in areas related to surgical training, intra-operative performance augmentation, robotic surgery and telecoaching. We use various methodologies, including artificial intelligence, AR/VR modelling, and simulation-based training to improve patient outcomes through surgical excellence.
  • Introduced MasTER, a data-intensive triage dashboard with a user-friendly human interface to enable fast patient dispatch in mass-casualty incidents by leveraging PPO-based deep reinforcement learning and large language models.

Ubisoft - Intern Gameplay Programmer

September 2022 - March 2023

Ubisoft is a leading video game company that develops and publishes popular game franchises such as Assassin's Creed, Far Cry, and Tom Clancy's Rainbow Six. Chengdu Studio's representative work is the Rabbids series.
  • Researched reinforcement learning (DQN, DDPG) on non-player character actions, behaviours, and interactions.
  • Assassin’s Creed Mirage downloadable contents (DLCs), excelling in C# and Unity3D and performance optimization.
  • Achieved notable improvements in DLC performance and functionality, streamlined project workflows with Perforce and Confluence, and successfully delivered high-quality content.

Academic

Dynamic Graphics Project - Graduate Student

January 2024 - April 2024

DocHub: Facilitating Comprehension of Documents via Structured Sensemaking with Large Language Models
  • Proposed DocHub, a LLM-based interactive system that identifies and visualizes crucial data and their interconnections within documents as node-link diagrams.
  • Offered an interactive interface allowing users to modify these visualizations for tailored insights and to pose detailed, context-specific queries for deeper understanding.
  • Featured a non-linear abstraction framework to adeptly handle and streamline the complexity of information presented.

Computational Social Science Lab - Graduate Student

September 2023 - December 2023

Who’s Hated: Detecting and Analyzing the Entities Targeted by Hateful Memes
  • Presented a pretrained language model-based framework to detect and reason about entities targeted by hateful memes.
  • Provided insight into why certain groups are more susceptible to becoming targets of hateful memes.
  • Proposed a specific preventive measure to curb the spread of hateful memes.

State Key Laboratory of Virtual Reality Technology and Systems - Researcher

February 2023 - June 2023

Efficient object manipulation is critical to VR interaction, and hands-free is a method worth discussing. We introduce a hands-free gaze-dominant manipulating pipeline that significantly outperforms the current state-of-the-art methods.
  • We proposed a hands-free object manipulation method based on gaze-dominant interaction, which significantly outperforms the current state-of-the-art gaze-based hands-free object manipulation method.
  • We designed a novel user study, facilitating a quantitative evaluation of the efficiency of the proposed method.

XDiscovery Lab, Dartmouth College - Intern Researcher

May 2022 - September 2022

We work on a wide spectrum of topics in human-computer interaction (HCI) with a specific emphasis on developing post-touchscreen hardware and software interactive systems, including new input/output technologies, interaction techniques, and novel augmentations for mobile and wearable devices.
  • Devised a novel text entry method that composes scattered keywords into a natural and clear sentence, which may help exaggerate the importance of human factors in studying natural language processing by, in this particular project, observing how people consider keywords.
  • Designed and developed a keyword extractor using BERT from huggingface.io.
  • Retrained the model based on the prompt-based approach to give three different semantic candidate sentences.
  • Developed a web application to enable more people to participate in our user study. Designed and implemented the UI with React framework. Stored data in MongoDB and used ExpressJS as the backend framework.

State Key Laboratory of Virtual Reality Technology and Systems - Researcher

September 2021 - February 2022

I led the team to devise CrossKeys, a novel and efficient text entry technique for virtual reality (VR) using a single controller via wrist rotation, which unprecedentedly employs the three-dimensional space a virtual environment can provide and outperforms the state-of-the-art method.
  • Led the team to devise CrossKeys, a novel and efficient text entry technique for virtual reality (VR) using a single controller via wrist rotation, which unprecedentedly employs the three-dimensional space a virtual environment can provide and outperforms the state-of-the-art method.
  • Implemented responsive components, auto-completing prediction algorithm, user interface design, ergonomics-mathematical deduction, and 3D modeling.

State Key Laboratory of Software Development Environment - Intern Researcher

March 2021 - December 2021

I developed Cross-Attention ReID, a state-of-the-art approach to realizing pedestrians' re-identification based on training with large-scale datasets generated by single-channeled IR cameras and three-channeled RGB cameras.
  • Developed Cross-Attention ReID, a state-of-the-art approach to realizing pedestrians’ re-identification based on training with large-scale datasets generated by single-channeled IR cameras and three-channeled RGB cameras.
  • Surveyed literature and applied existing theories to code with high performance and robustness.
  • Conducted quantitative analysis and results assessment with datasets like SYSU-MM01 and RegDB.

Services

CSC165: Mathematical Expression and Reasoning for Computer Science

Teaching Assistant

University of Toronto, January 2024 - April 2024

Skills

Note: I think this section is silly, but every company seems to require one.

NumPy
5 / 5
Unity3D
5 / 5
C#
5 / 5
Python
5 / 5
Git
4 / 5
AWS Services
4 / 5
Google Firebase
4 / 5
React
4 / 5
React Native
4 / 5
Linux/Unix
4 / 5
Matplotlib
4 / 5
PyTorch
4 / 5
Scikit-Learn
4 / 5
C++
4 / 5
Java
4 / 5
JavaScript
4 / 5
Apache Kafka
3 / 5
Apache Spark
3 / 5
Docker
3 / 5
Flask
3 / 5
Kubernetes
3 / 5
SciPy
3 / 5
Scala
3 / 5
TypeScript
3 / 5
PostgreSQL
2 / 5
SQL
2 / 5