01
Computational Optics
Co-designing optical modulation, sensor architecture, calibration, and reconstruction for compact high-performance imaging systems.
Computational Optics / Optical Computing / Computational Spectroscopy
Yuzhe Zhang
PhD student in Electrical and Computer Engineering at The University of Hong Kong, researching computational optics, optical computing, high-dimensional spectral sensing, and AI-driven image restoration.
HKU ECE
PhD, Computational Optics
Nature / CVPR
Selected publications
2 Patents
Granted spectral imaging IP
CAST 2025
Young Talent Program
Profile
Research Bio
I am a PhD student in Electrical and Computer Engineering at The University of Hong Kong, with a research focus on computational optics. My broader research interests include computational imaging, optical computing, computational spectroscopy, hyperspectral sensing, and AI-based visual reconstruction.
Before joining HKU, I received my M.Eng. in Information and Communication Engineering from Beijing Institute of Technology, where I worked with the Computational Imaging and Sensing Lab on hyperspectral image sensors, metasurface spectral imaging, and reconstruction algorithms. I also worked at SenseTime Research on diffusion-based image generation, restoration, and mobile deployment.
Computational Optics
Optical Computing
Computational Spectroscopy
Hyperspectral Imaging
Diffusion Restoration
Scientific Computing
Research Direction
Optics, Sensors, and Algorithms Designed Together
My work connects light modulation, compact optical hardware, inverse reconstruction, and vision foundation models for high-dimensional visual information acquisition.
02
Optical Computing
Exploring optical-domain encoding and computation to improve sensing efficiency, bandwidth, and physical interpretability.
03
Computational Spectroscopy
Building spectral sensors and reconstruction models for broadband, high-resolution, and wide-wavelength measurements.
Optical AI
Light Modulation
Spectral Sensors
Inverse Models
Diffusion Vision
Selected Work
Projects Across Optical Hardware and AI Reconstruction
Broadband Hyperspectral Image Sensor With High Spatio-Temporal Resolution
Proposed an on-chip hyperspectral imaging architecture based on broadband light modulation, with compact sensors covering 400-1000 nm and 400-1700 nm.
Diffusion-Based Blind Text Image Super-Resolution
Designed DiffTSR with image and text diffusion models to recover severely degraded text images with more accurate structures and realistic visual details.
Meta-Attention Spectral Reconstruction With Snapshot NIR Metasurface
Developed deep-unfolding reconstruction with multidimensional attention for low-crosstalk spectral imaging from compact metasurface measurements.
Broadband Miniature Spectrometer
Led student-side research on broadband miniature spectrometers using light modulation, quantum dots, filter arrays, and compressive sensing.
Publications
Selected Research Output
Authors marked with dagger contributed equally. Asterisk indicates corresponding author.
High-speed spectral imaging via multispectral pulse illumination and temporal-spectral decoupling
Jinxuan Wu, Daoyu Li, Jiajun Zhao, Hanwen Xu, Yuzhe Zhang, Liheng Bian
Optics Letters, 2025, vol. 50, no. 3, pp. 972-975.
A broadband hyperspectral image sensor with high spatio-temporal resolution
Liheng Bian†*, Zhen Wang†, Yuzhe Zhang†, et al.
Nature, 2024. Highlighted by Nature Photonics.
Meta-attention network based spectral reconstruction with snapshot near-infrared metasurface
Haoyang He†, Yuzhe Zhang†, Yujie Shao, et al.
Advanced Materials, 2024.
Diffusion-based Blind Text Image Super-Resolution
Yuzhe Zhang, Jiawei Zhang, Hao Li, et al.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
Deep calibration for broadband multispectral filter array
Yuzhe Zhang, Shuai Wang, Qiang Jiao, et al.
Oral Presentation, SPIE Photonics Asia, 2023.
Experience
Education and Research Experience
2025-2029
The University of Hong Kong
Doctor of Philosophy, Electrical and Computer Engineering. Research: Computational Optics.
2023-2025
SenseTime Research
Computer vision algorithm researcher in intelligent imaging; worked on ISP, diffusion models, transformers, face restoration, text image super-resolution, and mobile deployment. Received Outstanding Intern recognition.
2021-2024
Beijing Institute of Technology
Master of Engineering, Information and Communication Engineering. National Scholarship, Top 1%.
2021-2022
Huawei Technologies Co., Ltd.
Student team leader for broadband miniature spectrometer research, focusing on light modulation, quantum dots, spectral filter arrays, and compressive sensing.
Credentials
Honors, Patents, and Technical Stack
Selected Honors
- CAST Young Talent Program for PhD Students, 2025
- Excellent Master's Thesis of the Chinese Institute of Electronics, 2025
- Municipal Outstanding Graduate, Beijing, 2024
- National Scholarship for Graduate Students, 2023
Granted Patents
- CN115014522B, Integrated Computational Spectral Imaging Method and Device
- CN115112236B, Multispectral Sensor Calibration Method and Device
Technical Stack
- Computational optics, computational spectroscopy, and spectral filter arrays
- ISP, compressive sensing, calibration, and reconstruction algorithms
- Diffusion models, Transformer, ViT, DiT, VAE, pretraining, and finetuning
- Mobile model deployment and efficiency optimization
Contact