Click anywhere to close

Yi Chen

PhD Candidate in Pattern Recognition and Intelligent Systems

Yi Chen

About Me

I am currently a PhD student jointly affiliated with the State Key Laboratory of Multimodal Artificial Intelligence Systems at the Institution of Automation, Chinese Academy of Sciences and Zhongguancun Academy, under the supervision of Prof. Cheng-Lin Liu.

I received the B.E. degree from the School of Space Science and Technology at Xidian University in 2021. I earned my Master's degree in Electronic Information from the National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences in 2024, also advised by Prof. Cheng-Lin Liu.

State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS)
Zhongguancun Academy
Pattern Analysis and Learning Group (PAL)
Beijing 100190, China
yi.chen@nlpr.ia.ac.cn

Research Focus

My Ph.D. research focuses on artificial intelligence and deep learning methods, particularly at the intersection of large language models and adjuvant science (Ai for Adjuvant). Additionally, I investigate fundamental theories of multimodal large language models, including reliable reasoning and inference acceleration.

My work bridges AI theory with practical applications in scientific domains.

Research Interests

  • Recognition and generation methods for Chinese online text lines
  • Fundamental theories and applications of large language models and multimodal foundation models
  • Practical applications of AI methodologies in scientific domains (Ai for Science), particularly in biology, chemistry, and materials science research

If my research focus or interests align with your initiatives, please feel free to contact me via email. I look forward to potential collaborations!

Community Contributions

Contributed Materials Chemistry Case to PaddleScience

Successfully deployed and integrated the widely-used Crystal Graph CNN (CGCNN) model in the PaddleScience toolkit PaddleScience.

  • Complete implementation of crystal structure data preprocessing, graph network construction, training and prediction pipeline
  • Wrote detailed case documentation and user guide
  • Code merged into the official main repository [[PR #977]]
  • Became an official PaddleScience materials chemistry application case [[Case Documentation]]
PaddlePaddle PaddleScience Python PyG/Graph Neural Networks Materials Informatics

Academic Activities

The Fortieth AAAI Conference on Artificial Intelligence (AAAI 2026) 2025
Main Track Program Committee

Preprints

(*: equal contribution)

  • An Efficient Strategy for Data-constrained Machine Learning in Materials Science
    ChunTing Shao*, Yi Chen*, ShanMan Song, PeiPei Yang, QingBo Yan, Jian Xu, Gang Su
    Coming Soon!
  • The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM
    Shibo Gao, Peipei Yang, Haiyang Guo, Yangyang Liu, Yi Chen, Shuai Li, Han Zhu, Jian Xu, Xu-Yao Zhang, Linlin Huang
    arXiv
  • MeteorPred: A Meteorological Multimodal Large Model and Dataset for Severe Weather Event Prediction
    Shuo Tang, Jian Xu, Jiadong Zhang, Yi Chen, Qizhao Jin, Lingdong Shen, Cheng-Lin Liu, Shiming Xiang
    arXiv
  • The Hitchhiker's Guide to Scientific Agents: A Journey Through the Cosmos of Research Automation
    Xinming Wang, Aslan Feng, Jian Xu, Yi Chen, Haiyang Guo, Fei Zhu, Minsi Ren, Yuanqi Shao, Hongzhu Yi, Hongming Yang, Winston Hu, Tailin Wu, Xuyao Zhang, Cheng-Lin Liu
    TechRxiv GitHub
  • Advancing Adjuvant Research with MLLMs: An Open-Ended Benchmark and Formal Framework
    Yi Chen*, Yu Zhang*, Jian Xu, Xu-Yao Zhang, Hua Yue, Xinming Wang, Zequan Lyu, Wei Wei, Cheng-Lin Liu
    arXiv (Coming Soon)
  • Sparsity Meets Similarity: Leveraging Long-Tail Distribution for Dynamic Optimized Token Representation in Multimodal Large Language Models
    Yi Chen*, Gao-Tong Yu*, Jian Xu
    arXiv
  • VAGU & GtS: LLM-Based Benchmark and Framework for Joint Video Anomaly Grounding and Understanding
    Shibo Gao, Peipei Yang, Yi Chen, Han Zhu, Yangyang Liu, Wenxin Zhang, Linlin Huang
    arXiv

Publications

(*: equal contribution)

  • Recoverable Compression: A Multimodal Vision Token Recovery Mechanism Guided by Text Information
    Yi Chen, Jian Xu, Xu-Yao Zhang, Wen-Zhuo Liu, Yang-Yang Liu, Cheng-Lin Liu
    The 39th Annual AAAI Conference on Artificial Intelligence (AAAI 2025), 2025
    Link arXiv Code
  • Decoupling Layout from Glyph in Online Chinese Handwriting Generation
    Min-Si Ren, Yan-Ming Zhang, Yi Chen
    The 13th International Conference on Learning Representations (ICLR 2025), 2025
    arXiv Code
  • Recognition of Online Handwritten Chinese Texts in Any Writing Direction via Stroke Classification Based Over-Segmentation
    Yi Chen, Heng Zhang, Min-Si Ren, Cheng-Lin Liu
    The 27th International Conference on Pattern Recognition (ICPR 2024), 2024
    Link
  • Context-Aware Confidence Estimation for Rejection in Handwritten Chinese Text Recognition
    Yang-Yang Liu, Yi Chen, Fei Yin, Cheng-Lin Liu
    The 18th International Conference on Document Analysis and Recognition (ICDAR 2024), 2024
    Link
  • Improved Learning for Online Handwritten Chinese Text Recognition with Convolutional Prototype Network
    Yi Chen, Heng Zhang, Cheng-Lin Liu
    The 17th International Conference on Document Analysis and Recognition (ICDAR 2023), 2023
    Link

© 2023 Yi Chen | Academic Homepage

State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS) | Zhongguancun Academy

```