Renly Xinhai Hou

prof_pic.jpg

Xinhai Hou (Renly)

PhD Candidate @ UofM

Email: xinhaih@umich.edu

[Curriculum Vitae]

I am currently a Ph.D. candidate in Bioinformatics and Scientific Computing at the University of Michigan, Ann Arbor, where I am co-advised by Dr. Todd Hollon and Dr. Brian Athey. I hold an M.S. in Bioinformatics from the University of Michigan and a B.S. in Statistics from the Chinese University of Hong Kong, Shenzhen.

My research focuses on machine learning for computer vision (multimodal understanding and visual reasoning), with a strong focus on training foundation models and developing LLM-based agentic systems. I apply these methods to complex real-world domains such as medical imaging and clinical decision support, creating models that reason more reliably and operate effectively in high-stakes environments.

Before starting my Ph.D. program, I worked as a machine learning engineer intern at Tencent under the supervision of Dr. Pengfei Xiong, where I contributed to projects on title generation and video tagging.

News

May 19, 2025 I start my internship as Applied Scientist at Amazon in Seattle.
Nov 14, 2024 Our paper Foundation models for fast, label-free detection of glioma infiltration was accepted in Nature 2024.
Feb 27, 2023 Our paper Hierarchical Discriminative (HiDisc) Learning Improves Visual Representations of Biomedical Microscopy was accepted and got highlight in CVPR 2023.
Sep 25, 2022 Our paper OpenSRH: Optimizing Brain Tumor Surgery Using Intraoperative Stimulated Raman Histology was accepted in NeurIPS 2022 Dataset and Benchmark track.

Selected Publications

  1. arXiv
    CodeV: Code with Images for Faithful Visual Reasoning via Tool-Aware Policy Optimization
    Xinhai Hou, Shaoyuan Xu, Manan Biyani, and 4 more authors
    2025
  2. Nature
    Foundation models for fast, label-free detection of glioma infiltration
    Akhil Kondepudi, Melike Pekmezci, Xinhai Hou, and 8 more authors
    Nature, 2024
  3. NeurIPS AIM-FM
    A self-supervised framework for learning whole slide representations
    Xinhai Hou, Cheng Jiang, Akhil Kondepudi, and 4 more authors
    Neural Information Processing Systems Workshop on Advancements In Medical Foundation Models, 2024
  4. CVPR
    Hierarchical discriminative learning improves visual representations of biomedical microscopy
    Cheng Jiang*, Xinhai Hou*, Akhil Kondepudi, and 5 more authors
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023
  5. NeurIPS D&B
    OpenSRH: optimizing brain tumor surgery using intraoperative stimulated Raman histology
    Cheng Jiang*, Asadur Chowdury*, Xinhai Hou*, and 7 more authors
    Advances in neural information processing systems, 2022