Title：AI4Astro: Exploring Star Formation and the ISM through Artificial Intelligence

Speaker：Duo Xu 许铎, University of Virginia

Time: 9:30am-10:30am, June 20th (Thursday)

Tencent Meeting：105-982-198password: 0620

Location: Small conference room 3rd floor

Abstract：

The field of astronomy is experiencing a profound shift driven by machine learning, particularly deep learning, which enables efficient processing of vast datasets, surpassing human capabilities in complex data analysis. In this talk, I will showcase the diverse capabilities of AI in astronomy, focusing on tasks related to star formation and the interstellar medium. I will present our AI methodologies, including discriminative and generative models, applied to tasks such as segmenting stellar feedback structures from molecular line data cubes, inferring outflow properties through regression analysis, and determining magnetic field directions. Additionally, we tackle challenging tasks involving the inference of physical quantities like volume density, interstellar radiation field, and magnetic field strength. Through this presentation, aim to highlight the transformative impact of AI on data analysis in astronomy and encourage exploration of these cutting-edge technologies. Let's embark on the AI journey together!

CV：Dr. Duo Xu is currently a VICO (VIRGINIA Initiative on Cosmic Origins) fellow at the University of Virginia. Starting this fall, Dr. Xu will be joining CITA (Canadian Institute for Theoretical Astrophysics) as a CITA Postdoctoral Fellow and a Schmidt AI in Science Postdoctoral Fellow at the University of Toronto. Dr. Xu completed his PhD at the University of Texas at Austin in 2021 and obtained his bachelor's degree from Nanjing University and his master's degree from the University of Chinese Academy of Sciences, National Observatory of China. His research focuses on star formation and the interstellar medium, integrating magnetohydrodynamic simulations and observational studies. Additionally, he has a keen interest in applying machine learning techniques to address challenges in astrophysics.