About me

贺图：绿荫大道尽头是北交大主楼，白发角色张开双臂站在光里，白鸽环绕飞舞
贺图 · Hewkick 赠 · Illustrated by @萤火

Email GitHub Google Scholar

自我介绍 Biography

  • 我叫翟王宇轩，是北京交通大学计算机科学与技术学院人工智能专业的本科生，现已保研至清华大学深圳国际研究生院。目前，我的主要研究领域是 音频大模型🎶, 特别是其中的语音合成 (Text-To-Speech, TTS) 与音乐生成方向。
  • My name is Wangyuxuan Zhai, an undergraduate student majoring in Artificial Intelligence at the School of Computer Science and Technology, Beijing Jiaotong University. I have been admitted to the master's program at Tsinghua Shenzhen International Graduate School via recommendation. Currently, my main research area is Audio Language Models 🎶, with a particular focus on Text-To-Speech (TTS) and music generation .

研究兴趣 Research Interests

  • 目前，我的主要研究方向是 语音合成和音乐生成🎶；
  • Currently, my main research interests are Text-To-Speech (TTS) and Music Generation🎶.

杂项 Misc

  • 除了研究之外，我还是一名音乐爱好者。我会一点手风琴🪗，中央音乐学院业余八级水平。
    我也会制作电子音乐🎵，自娱自乐足矣。
  • 我从小学开始就对术力口（Vocaloid）与音乐节奏游戏（音游）感兴趣，这深刻地影响了我对于未来专业方向的选择。
  • Outside of research, I am also a music lover. I play a bit of accordion 🪗 and have reached amateur Grade 8 in the Central Conservatory of Music system. I also make electronic music 🎵, mostly just for fun.
  • I have been fascinated by Vocaloid and rhythm games since elementary school, and they have had a deep influence on the academic path I chose for myself.
「不诱于誉，不恐于诽，率道而行，端然正己。」

CV

Education

  1. Tsinghua Shenzhen International Graduate School
    清华大学深圳国际研究生院

    2027.09 — 2029.07

    硕士研究生 · Master's

  2. Beijing Jiaotong University
    北京交通大学

    2023.09 — 2027.07

    本科生 · Undergraduate

Experiences

  1. ModelBest
    面壁智能

    2026.05 - Present

    Audio Large Model Algorithm Intern 音频大模型算法实习生

  2. Human-Computer Speech Interaction Lab at Tsinghua University (THU HCSI)
    清华大学深圳国际研究生院人机语音交互实验室

    2026.04 - Present

    Text-To-Speech & Music Generation 语音合成 & 音乐生成

  3. Institute for AI Industry Research, Tsinghua University (THU AIR)
    清华大学智能产业研究院

    2025.09 - 2026.04

    LLM + Reinforcement Learning 大模型 + 强化学习

  4. BJTU NLP Group
    北交自然语言处理团队

    2024.09 - 2025.05

    LLM + Geography 大模型 + 地理信息

Publications

  1. Please refer to my Google Scholar.

    请参考我的 谷歌学术.

  2. X. Wang*, Z. Kang*, W. Zhai*, X. Lou, Y. Lai, Z. Wang, Y. Wang, K. Huang, Y. Wang, P. Li, Y. Liu. (2025). MUCAR: Benchmarking Multilingual Cross-Modal Ambiguity Resolution for Multimodal Large Language Models. EMNLP 2025 Main Conference. [Paper]

  3. L. Yuan*, F. Mo*, K. Huang, W. Wang, W. Zhai, X. Zhu, Y. Li, J. Xu, J.-Y. Nie. (2025). OmniGeo: Towards a Multimodal Large Language Models for Geospatial Artificial Intelligence. arXiv Preprint.[Paper]

AWARDS

  1. 北京交通大学 2023-2024 年度 国家奖学金

    National Scholarship, Beijing Jiaotong University, 2023-2024 Academic Year

  2. 北京交通大学 2024-2025 年度 国家奖学金

    National Scholarship, Beijing Jiaotong University, 2024-2025 Academic Year

Others

If you'd like to engage in a discussion or collaborate, feel free to contact me via email at any time! 🥰🥰🥰

✉️ zhaiwangyuxuan [at] bjtu.edu.cn

Here's my friend's blog. Feel free to drop by and have a look! 下面是我朋友的博客入口，欢迎串门！！！

朋友 Friends

  • Luyi Lin / speedlast

    致力于基础模型、AI infra、硬件感知算法以及 AI4Math 领域研究。
    Working on foundation models, AI infrastructure and hardware-aware algorithms, and AI for math.

  • Jian Yan (Cris Yan)

    致力于推动人工智能普惠通用，目前专注于推理加速方向。
    Committed to the universal use of AI. Now focusing on inference acceleration.

  • Yukai Wu

    专注于大模型/多模态大模型驱动的后训练与智能体系统研究。
    Focusing on post training / agent system powered by LLM/MLLM.

  • Hewkick

    同样也在音频方向工作；我们会一起合作电子音乐。 Bilibili 链接 | 网易云音乐 链接
    Also works on audio; we collaborate on electronic music together — Bilibili Link | NetEase Cloud Music Link.