Homepage: https://jarvisustc.github.io/

I am a fourth-year Ph.D. student in the joint program between the University of Science and Technology of China (USTC) and Microsoft Research Asia (MSRA), co-supervised by Prof. Qiang Huo at MSRA and Prof. Jun Du at USTC. My Ph.D. research focuses on Document Intelligence (including OCR, document layout analysis, and document understanding) and Large Language Models (including MLLM, Agent, and RAG). Prior to this, I received my B.S. degree from the School of the Gifted Young (a.k.a. 少年班) at the University of Science and Technology of China in 2021, majoring in Computer Science.

During my Ph.D. studies, I gained valuable industry experience through internships at MSRA, DeepSeek, and ByteDance. At DeepSeek, I contributed to the development of DeepSeek VL2, DeepSeek V3, and DeepSeek R1. My internship at MSRA involved working on the Microsoft OneOCR project and the Microsoft Document Intelligence project under the guidance of Researcher Qiang Huo and Lei Sun. Most recently, I began an internship with the ByteDance Seed team, where I am working on LLM/MLLM Agent projects. I have published over 10 papers at top-tier international AI journals and conferences.

I am currently seeking full-time job opportunities. If you are interested in my resume, please feel free to email me at [email protected]. I am currently based in Beijing, China. If you would like to have a coffee chat, please feel free to reach out! ☕😊✨

• 2025.04: 🎉 Thrilled to kick off my new internship with the ByteDance Seed Team!
• 2025.04: 🔥 Our latest work on VLM robustness, "Safeguarding Vision-Language Models: Mitigating Vulnerabilities to Gaussian Noise in Perturbation-based Attacks," has been released on Arxiv. We've open-sourced the Robust-VLGuard dataset and DiffPure-VLM defense.
• 2025.03: 🎉 Our paper "UniHDSA: A Unified Relation Prediction Approach for Hierarchical Document Structure Analysis" has been accepted by Pattern Recognition Journal!
• 2024.12: 💻 We've launched a new GitHub project: Awesome-Multimodal-RAG! Check out the latest in multimodal RAG and contribute!
• 2024.12: 🤝 We're excited to have contributed to DeepSeek-VL2, an advanced Vision-Language Model with strong performance and fewer parameters.
• 2024.08-09: 🗣️ Presented DLAFormer and DRFormer at ICDAR in Athens! Photos can be found here. A memorable experience meeting colleagues and exploring the city.
• 2024.08: ✍️ The complete version of DLAFormer, titled "UniHDSA: A Unified Relation Prediction Approach for Hierarchical Document Structure Analysis", has been submitted to Pattern Recognition Journal.
• 2024.07: 🎉 Our Detect-Order-Construct have been accepted by Pattern Recognition!
• 2024.06: 🗣️ Our DLAFormer, UniVIE, and DRFormer selected for oral presentation at ICDAR 2024!

• 2025.4-Now: Research Intern, Seed Team, ByteDance , Beijing, China.
• 2024.09-2025.03: Research Intern, Multimodal Interaction Group, Microsoft Research Asia , Beijing, China.
• 2024.06-2024.08: AGI Research Intern, Multimodal LLM Team, DeepSeek , Beijing, China.
• 2020.09-2024.05: Research Intern, Multimodal Interaction Group, Microsoft Research Asia , Beijing, China.

• 2021.09-2026.6: Ph.D. in Information and Communication Engineering, University of Science and Technology of China, Hefei, Anhui, China.
• 2017.09-2021.06: B.S. in the School of the Gifted Young (major in Computer Science), University of Science and Technology of China, Hefei, Anhui, China.