Biography
I am Peiyang Ni (倪培洋), a Master's student in Computer Science at the University of Electronic Science and Technology of China (UESTC), advised by Prof. Ping Hu and Prof. Xiaofeng Zhu. I received my B.Eng. degree in Computer Science from Guilin University of Electronic Technology (GUET).
My research focuses on image generation with diffusion and flow models, and physics-informed generative modeling for high-fidelity 3D perception. I am particularly interested in LiDAR sensing and measurement formation, as well as realistic generation and restoration under adverse weather.
Publications
Structure-to-Intensity Diffusion for Adverse-Weather LiDAR Generation
CVPR 2026
A physics-grounded diffusion framework that reconstructs LiDAR structure before denoising reflectance intensity, supported by real-prior adverse-weather simulation.
A2CFlow: Range-Aware Rectified Flow for Efficient Adverse-Weather LiDAR Restoration
Under Review at IEEE Robotics and Automation Letters (RA-L)
A conditional rectified-flow framework that restores clean range-and-reflectance LiDAR scans from a single adverse-weather observation, with range-aware flow scheduling and efficient few-step distillation.
Experience
Education
University of Electronic Science and Technology of China
2024–PresentM.Eng. in Computer Science · Chengdu, China
GPA: 3.98 / 4.00 · Advisors: Prof. Ping Hu and Prof. Xiaofeng Zhu
Guilin University of Electronic Technology
2020–2024B.Eng. in Computer Science · Guilin, China
Average: 89.8 / 100
Awards
- National Scholarship for Undergraduate Students (0.2%), Ministry of Education of China
- CASC Scholarship (Third Prize), China Aerospace Science and Technology Corporation
- Guangxi Zhuang Autonomous Region People's Government Scholarship
- Outstanding Graduate of Guangxi Zhuang Autonomous Region
- Outstanding Graduate Student, University of Electronic Science and Technology of China
Academic Service
Conference Reviewing
- PRCV
- VCIP
Journal Reviewing
- IEEE Robotics and Automation Letters (RA-L)