Mengyu Zhao

PhD student @ Rutgers University

mypic.jpg

Hi! I’m Mengyu Zhao, a PhD student in Electrical and Computer Engineering at Rutgers University, where I am fortunate to be advised by Prof. Shirin Jalali. My research focuses on visual restoration using image priors and the development of efficient LLM/MLLM agents.

I am currently an intern at TikTok Inc., focusing on efficient vision encoders. Previously, I spent the summer of 2025 as an intern at Bloomberg, where I developed GUI-based LLM agents. I hold a BE from Xi’an Jiaotong University and an MS from Rutgers University.

I’m always open to technical discussions and collaboration. I also help organize a reading group dedicated to deep learning for inverse problems.

Email: mengyu.zhao [at] rutgers [dot] edu

selected publications

  1. ICASSP
    Saturation in Snapshot Compressive Imaging
    Mengyu Zhao, and Shirin Jalali
    International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2026
    arXiv
  2. SIIMS
    Theoretical Characterization of Effect of Masks in Snapshot Compressive Imaging
    Mengyu Zhao, and Shirin Jalali
    SIAM Journal on Imaging Sciences (SIIMS), 2025
    arXiv
  3. NeurIPS
    Untrained Neural Nets for Snapshot Compressive Imaging: Theory and Algorithms
    Mengyu Zhao, Xi Chen, Xin Yuan, and Shirin Jalali
    Advances in Neural Information Processing Systems (NeurIPS), 2024
    arXiv Code
  4. Allerton
    Theoretical Analysis of Binary Masks in Snapshot Compressive Imaging Systems
    Mengyu Zhao, and Shirin Jalali
    2023 59th Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2023
    arXiv