【6.2; Seminar】2026年粒子天体物理全国重点实验室系列学术报告 #19: Physics-informed AI for Astronomical Observation：Expanding the Scientific boundary

时间/Time: 2026年6月2日周二上午9:40 / Tuesday, June 2, 2026, 9:40 a.m. 

地点/Location: 多学科122会议室 / 122 Meeting Room, Multidisciplinary Building

报告人/Speaker: 郭钰铎（清华大学） / Dr. Yuduo GUO (Tsinghua University)

题目/Title: Physics-informed AI for Astronomical Observation：Expanding the Scientific boundary

主持人/Host: 冯骅 研究员 / Prof. Hua FENG

摘要/Abstract:

Science begins with measurement. While visual technologies driven by physical optics continually expand scientific frontiers, further scaling physical hardware now faces diminishing returns. This talk will focus on physics-informed AI for Astronomy and present a science-driven framework that integrates AI for computational optics with physics-inspired neural networks, with the goal of moving beyond the conventional dichotomy between physical instrumentation and data-driven intelligence. Centered on the scientific objectives of next-generation observation and measurement, the talk will examine three tightly connected directions: high-dimensional light-field acquisition, high-precision correction of complex optical propagation, and high-fidelity reconstruction of ultra-weak photons. Through these directions, I will discuss several fundamental challenges in astronomical observation, including optical aberrations, atmospheric turbulence, and background photon noise, and show how the integration of computational optics with physics-informed AI can provide an alternative pathway to overcoming detection limits imposed by physical bottlenecks. More broadly, this line of research points toward a new paradigm of intelligent scientific measurement, with potential impact extending beyond astrophysics to other fundamental disciplines such as the life sciences.

个人简介/Biography:

Yuduo Guo is the Assistant Professor of College of AI at Tsinghua University. His research focuses on Computational Optics and Artificial Intelligence. He has published in journals such as Science, Nature, Cell, and Nature Photonics, and holds over ten national invention patents. His research on intelligent Meta-imaging and Digital adaptive optics has received academic recognition, including the 2023 Frontiers of Science Award at ICBS, the 2022 Top Ten Advances in Optics in China, and the Tsinghua Outstanding Doctoral Dissertation Award. Related research outcomes have been translated into practical applications at institutions such as the National Astronomical Observatories.