My name is Wang Zeyu, and I am currently a third-year undergraduate student majoring in Computer Science and Technology at Taiyuan University of Technology (School of Computer Science and Technology, School of Big Data; Project 211 university in China). I have had a strong passion for computer science since childhood, teaching myself technologies such as Python and Linux from a young age. I also developed and deployed my personal website 1800+ days. In my spare time, I contribute to opensource projects. My GitHub is: https://github.com/moyitech

During my undergraduate studies, I joined the Multimodal Human-Robot Interaction (MIHRI) Lab led by Associate Professor (Yidi Li). There, I participated in a research project on audiovisual speaker tracking, where I implemented a high-precision tracking algorithm with pixel-level accuracy (around 3-pixel error). The related work has been accepted by ICASSP 2025, where I am listed as the second student author.

In addition, I have a strong interest in large language models (LLMs) and reinforcement learning (RL), and have completed three industry internships. I am currently working at Baidu as a research intern focused on foundation model algorithms.

Publications

• Yidi Li, Wenkai Zhao, Zeyu Wang, Zhenhuan Xu, Bin Ren, Nicu Sebe. Multi-Stage Multimodal Distillation for Audio-Visual Speaker Tracking, Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2025.

InternShip

May 2025 - Present

🏢 Baidu | E-commerce Search Innovation Technology Group

💼 NLP Algorithm Research Intern

Project: Construction of a Large Language Model for Named Entity Recognition in Baidu Search E-commerce Field

Background: The project aims to use reinforcement learning techniques to construct a named entity recognition language model for the e-commerce search field by exploring unknown products, in order to optimize content retrieval in the e-commerce field.

🎉 Key Contributions:

Comming soon…

November 2024 - April 2025

🏢 SentimentTrader | Hangzhou R&D Center

💼 AI Agent Algorithm Intern

Project: Construction of Intent Recognition and Intelligent Customer Service Large Model in the Financial Domain

Background: This company is an information service provider in the financial sector. Currently, it is offering its clients AI Agent services that include investment advice.

🎉 Key Contributions:

🤖 A professional AI Agent has been constructed, with the following details:

1. RAG system: By using the analysis articles provided by professional analysts within the station, it offers accurate QA services to users.
2. Agent System: Based on the user’s intention, the GPT is used to perform continuous page operations for the backtesting tool.

December 2023 - June 2024

🏢 Strikingly | Hangzhou R&D Center

💼 RAG System Development Intern

Project: Build a multilingual and multi-functional RAG dialogue translation system.

Background: This company offers services globally, but the majority of its customer service staff are in English, and there are many technical terms in the conversations. Therefore, an AI Agent is needed to assist them in completing reliable translation tasks.

🎉 Key Contributions:

I mainly developed two AIGC applications based on GPT-4:

1. Customized Translation: Using the RAG (Retrieval-Augmented Generation) architecture concept, Milvus vector database and specific domain international vocabulary library, we achieved precise translation output for complex context conversations, with the results presented in structured JSON format.
2. Customer Service Assistance QA: By segmenting and integrating document resources, using Agentic-RAG for real-time search and matching of relevant text fragments, we automatically generated targeted responses along with source text links, effectively empowering the customer service team to quickly and accurately answer user questions.

Competitions

• 2023 National University Students’ Mathematical Modeling Competition | Coder | 🏆 National Second Prize
• 2023 China Mobile Wutong Cup Big Data Competition | Developer | 🏆 National Third Prize | Top10
• 2023 Alibaba Cloud Tianchi AgentBuilder Challenge | Team Leader & Developer | 🏆 NVIDIA Technology Award
• 2024 China University Students’ Intelligent Lighting and Intelligent Wearable Innovation Competition | Team Leader & Developer | 🏆 National Special Prize
• 2024 International Youth Artificial Intelligence Competition | Team Leader & Developer | 🏆 International First Prize
• 2024 Global Campus Artificial Intelligence Algorithm Elite Competition | Team Leader & Developer | 🏆 National Second Prize
• 2024 National University Students’ Career Planning Competition | 🏆 Top 1 of Taiyuan University of Technology

OpenSource Contribution

AI Opensource Learning Project

• Self-LLM(Open Source Large Model User Guide): https://github.com/datawhalechina/self-llm/ | Core Contributor | 15.5k stars 🌟
• Tiny-Universe | https://github.com/datawhalechina/tiny-universe | 2.9k stars 🌟

Mutual AI

We created this project to make the boring and difficult-to-understand algorithms of AI interesting, and to teach learners how to apply AI to real-life situations.

GitHub link :https://github.com/YinHan-Zhang/Mutual-AI