Research Projects
National Mathematical Modeling Competition
I participated in the three-day National Mathematical Modeling Competition, where our team designed an optimal planting scheme under specified constraints. As team leader, I was primarily responsible for constructing mathematical models and programming. Ultimately, we completed a research paper and presented our findings, earning a provincial third prize. This marked my first engineering practice applying abstract mathematical concepts to real-world problems.
Research: Graph Transformers and Sparse Attention Mechanisms
With support from the National Program for Cultivating Elite Undergraduate Students, I had the opportunity to independently conduct research in graph learning. Collaborating with an undergraduate from another university, we explored integrating sparsification and attention mechanisms into Transformer-based models. By applying these methods to benchmark datasets, we achieved higher computational efficiency and improved prediction accuracy, demonstrating the potential of sparse attention for scalable graph learning.
Engneering Projects
Attempts in Smart Agriculture: AgriBulter
We designed an intelligent agriculture management system called AgriButler, which integrates AI, IoT, and computer vision technologies. The system relies on mobile robots and sensor networks for farmland monitoring and service management, with the goal of improving agricultural productivity, reducing costs, and advancing the modernization of agriculture. I am the primary responsible person for this project.
Collaborative Projects
Digital Human for Promoting Chinese Opera
This project leveraged voice synthesis, motion capture, and digital human technologies to create virtual characters capable of performing Chinese opera. The goal was to preserve and promote traditional Chinese opera culture through digital media. In this project, I am primarily responsible for pose extraction technology.
SmartLesson: AI-powered Teaching Preparation System
We developed an AI-assisted lesson preparation system based on large language models. The system can automatically generate teaching outlines, lesson plans, quizzes, and supporting materials, significantly reducing teachers鈥?workload while promoting personalized education. In this project, I was primarily responsible for technical analysis and showcasing results.
AI-powered Drug Target Research Platform
We built a research platform that applies AI algorithms and multi-dimensional data analysis to predict the efficacy of nucleic acid drugs and generate optimization suggestions. This platform aims to accelerate drug development while reducing laboratory costs. My main responsibility was technical analysis and presentation of research outcomes.
Design and Flight Testing of a Simple Quadcopter
This project comprehensively covers the entire design process of a quadcopter: from selecting basic components to final assembly and flight testing. Through this hands-on experience, I have gained engineering expertise in PCB development, hardware design, and modeling and analysis of flight propulsion systems.