Huayi's Personal Website

Huayi Chen

To be

a Machine Learning Engineer.

a Data Engineer.

a Data Scientist.

Huayi Chen

Hi, I want to be

a Machine Learning Engineer.

a Data Engineer.

a Data Scientist.

A passionate and hard-working programmer.

About Me

I am an MSc in Computer Science (AI Specialization) student at York University. My work focuses on Computer Vision and Deep Learning, with a strong background in technical research and application development.

Skills

Programming

Imagination

Team Work

Patience

Education

Academic

Master of Computer Science

2025-2027 / York University

Toronto, ON

Bachelor of Computer Science

2018-2022 / University of Manitoba

Winnipeg, MB

Work Experience

Professional

Research Assistant Intern

Sep 2025 - Dec 2025 / Fuzhou University

• Developed intelligent car and robotic arm systems based on the Huawei Atlas 200I DK A2 platform, implemented AI-driven functionalities, and optimized the workflows and standards of teaching experiments.
• Implemented intelligent car's keyboard control, autonomous driving (lane keeping + road sign detection), automatic parking; deployed ResNet18/YOLOv5 models for efficient trajectory tracking.
• Achieved robotic arm's basic control, garbage sorting, and gesture interaction.
• Collaborated with professors to enhance teaching documents and experiment protocols, streamlined operational procedures, reduced learning curves for students, and effectively elevated the overall teaching implementation effect.

Research Assistant (Seasonal)

Sep 2023 - Aug 2025 / Fuzhou University

Conducting research on Computer Vision, with a focus on video object detection and trajectory reconstruction. Published papers in IEEE ICIP '25, IEEE ICME '25, and PRCV '24.
Key Projects & Outcomes:
- (1) Occluded Video Object Detection (IEEE ICIP '25):
  - • Designed a spatiotemporal context-aware network to solve occlusion issues.
  - • Achieved SOTA performance on benchmark datasets (83.2% mAP on ImageNet VID, 63.7% on OVIS).
- (2) Graph-based Trajectory Reconstruction (PRCV '24):
  - • Integrated YOLOv6 and ResNet50 with a graph-based method to improve object association.
  - • Reduced reconstruction RMSE by ~10% and outperformed baselines (e.g., YOLOv6+GSI) by ~1.5%.
Tech Stack: Python, PyTorch, YOLOv6, ResNet, OpenCV.