Ozzy Chen Avatar

Hello, I'm Ozzy Chen

Student at UNNC | Mathematics & Applied Mathematics
Bridging Math Theory with Programming.

01. About Me

I am currently pursuing my degree in Mathematics & Applied Mathematics at the University of Nottingham Ningbo China (2024–2028).

My passion lies in the intersection of rigorous mathematical theory and practical algorithmic implementation. I focus on algorithm optimization and finding quantitative solutions to complex problems using C++ and Python.

"Mathematics is the poetry of logical ideas"
— Albert Einstein
Ozzy Chen Profile

02. Experience & Education

Research Assistant

July 2025 – Present
Shenzhen University of Advanced Technology (AI Center)

Computer Vision and Recognition Center (AI觉-知研究中心)

  • Conducting research on Large Visual Language Models (LVLMs), studying efficiency optimization.
  • Investigating reasoning problems in long-context image understanding.
  • Bridging sensory input with cognitive understanding using tools like Qwen2.5-VL.

University of Nottingham Ningbo China

2024 – 2028

B.Sc. Mathematics & Applied Mathematics

Relevant Coursework: Linear Algebra, Machine Learning, Deep Learning, Computer Vision.

03. Tech Stack

Languages

C++ Advanced
Python Proficient
MATLAB / LaTeX Academic

Libraries & Tools

PyTorch Qwen2.5-VL OpenCV NumPy Pandas Git Linux

Domains

  • ▹ Large Visual Language Models
  • ▹ Algorithm Design
  • ▹ Data Structures
  • ▹ Mathematical Modeling

04. Projects & Achievements

arXiv

TC-SSA: Token Compression via Semantic Slot Aggregation

Proposed a learnable token-budgeting mechanism for Gigapixel Whole-Slide Images. Reframed spatial reasoning as aggregation-based compression via gated assignment (Top-2 routing), achieving 76.78% accuracy on SlideChat and 120x efficiency gain.

PyTorch VLM WSI Analysis