CV

Ph.D. Candidate

ShanghaiTech University, Shanghai, China

• Computer Science and Technology
• Advisor Yuyao Zhang

Bachelor

Wuhan Univeristy of Technology, Wuhan, Hubei, China

• Computer Science and Technology

Solving Medical Inverse problem via Self-supervised learning.

• Adopted Implicit Neural Representation for CT and MRI reconstruction.
• Adopted diffusion model for MRI motion correction.

Self-supervised Image Denoising for Biomedical images.

• Proposed a self-supervised image denoising framework Noise2SR.
• Theoretical proved that a denoised image can be generated from a sub-sampled noisy image using whole noisy image supervision which is equivalent to using clean signal as supervision.
• Experiment results show that Noise2SR outperforms SOTA self-supervised denoising methods (Noise2Self, Neighbor2Neighbor) in real fluorescence microscopy and MRI image noise removal.

Zero-shot Learning for High Resolution Electron Microscopy (HREM).

• Proposed a zero-shot learning self-supervised denoising framework for noisy HREM images.
• Outperformed self-supervised SOTA zero-shot methods (Self2Self, FBI-Denoiser) in ultra-low SNR HREM image scenarios

Teaching Assistant

CS270B: Advanced Deep Image Processing, ShanghaiTech University