Zheyu Huang – Curriculum Vitae

Zheyu Huang Wenzhou, Zhejiang, China Email: brandhzysds@gmail.com | Phone: +86 1361662678 / +61 422845661


Professional Summary

Final-year Bachelor of Engineering (Honors) student majoring in Software Engineering at the University of Technology Sydney (UTS). Passionate about artificial intelligence, computer vision, and deep learning, with a strong focus on multimodal fusion and non-ideal supervision learning. Experienced in software development, object detection, and real-time image segmentation for high-precision infrared monitoring systems. Skilled in machine learning model optimization, data analysis, and software architecture.


Education

University of Technology Sydney (Joint Program) – Australia Bachelor of Engineering (Honors) – Software Engineering (08/2023 – Present)

Northeastern University (Joint Program) – China Bachelor of Computer Science (09/2021 – 06/2023)


Research & Projects

High Precision Long Range Infrared Object Monitoring System

(August 2024 – July 2025)

Developed a high-precision infrared monitoring system using the Sigma Siamese Mamba Network for detecting vehicles, pedestrians, and wildlife in various environments.

Designed and optimized modular architecture for high-resolution, long-distance infrared image segmentation.

Implemented a real-time inference pipeline integrating FLIR FC-608O thermal camera for robust dataset collection.

Conducted experiments on boundary blurring, noise reduction, and segmentation accuracy improvement.

Applied research findings to wildlife monitoring and environmental surveillance.

Comparative CLIP: Visual Classification with Filtered Comparative Descriptors

(August 2024 – September 2024)

Designed a zero-shot classification framework improving vision-language models like CLIP by integrating comparative reasoning and similarity filtering.

Enhanced semantic retrieval through GPT-4-generated comparative descriptors for visually similar categories.

Achieved superior generalization and interpretability on datasets such as ImageNet and CUB200-2011.

Improved classification accuracy by 10.5% on average over baseline models.


Professional Experience

Codnoy Technology – Developer Intern (August 2022 – October 2022)

Developed core backend functionalities using SpringBoot, designing a four-layer architecture and optimizing APIs.

Designed and implemented database schemas and table structures to enhance data storage and retrieval efficiency.

Provided technical support and guidance to junior developers, improving overall team collaboration.

Recognized as an "Outstanding Developer" for innovative performance optimizations.


Skills

Programming: Python, Java, C++, JavaScript, SQL

Frameworks & Tools: PyTorch, TensorFlow, Hugging Face Transformers, OpenCV, LaTeX, Linux

Expertise: Machine Learning, Deep Learning, Image Segmentation, Software Architecture, Data Analysis

Languages:  English and Mandarin