Xianyuan Liu

Academic Fellow in Multimodal AI (starting March 2026) @The University of Sheffield.

prof_pic.jpg

I am currently a Senior AI Research Engineer and will soon take up the role of Academic Fellow in Multimodal AI (from March 2026) in the School of Computer Science, University of Sheffield. I am also the Assistant Head of AI Research Engineering at the Centre for Machine Intelligence.

My research focuses on developing machine learning methods and tools for materials science and multimodal learning. In particular, I use multimodal AI and large language models to construct new datasets and benchmarks for materials research, and develop multimodal AI methods for improved materials property prediction using both computational and experimental data.

I am a co-founder and principal developer of PyKale, an open-source library that delivers accessible multimodal machine learning algorithms, which is featured in PyTorch Landscape for multimodal training.

Selected Publications

  1. Nature MI
    Towards deployment-centric multimodal AI beyond vision and language
    Xianyuan Liu, Jiayang Zhang, Shuo Zhou, and 45 more authors
    Nature Machine Intelligence, 2025
  2. Nature MI
    Mask-prior-guided denoising diffusion improves inverse protein folding
    Peizhen Bai, Filip Miljković, Xianyuan Liu, and 4 more authors
    Nature Machine Intelligence, 2025
  3. IEEE TCSVT
    First-person video domain adaptation with multi-scene cross-site datasets and attention-based methods
    Xianyuan Liu, Shuo Zhou, Tao Lei, and 3 more authors
    IEEE Transactions on Circuits and Systems for Video Technology, 2023
  4. CIKM
    PyKale: Knowledge-aware machine learning from multiple sources in Python
    Haiping Lu, Xianyuan Liu, Robert Turner, and 5 more authors
    In Proceedings of the 31st ACM International Conference on Information and Knowledge Management, 2022