معرفی
Anran Xu is a Researcher in the Department of Earth, Ocean and Atmospheric Sciences at the University of British Columbia (UBC), within the Faculty of Science. Their work focuses on applying machine learning techniques to solve inverse problems across multiple physics domains, including geophysics, cosmology, and biomedical imaging.
Education and Supervision: Anran completed their Master’s thesis under Prof. Qinya Liu at the University of Toronto (UofT), focusing on earthquake-related inverse problems. They also collaborated with Dr. Keir Rogers (UofT) on cosmology projects and with Luca Calatroni and Laure Blanc-Féraud at CNRS on fluorescence microscopy applications.
Research Interests: Central to their work is the development of test-time learning and neural network parameterizations to enhance geophysical inversion methods. Their research spans AI-driven solutions for parameter estimation in complex physical systems, emphasizing regularization techniques and neural field applications. They explore cross-disciplinary challenges in geophysics, cosmology, and microscopy through computational frameworks.
Publications: Their recent articles highlight contributions in leveraging CNNs for implicit regularization in DC resistivity inversions and reparameterizing geophysical inverse problems using neural networks. These works reflect a trend toward integrating modern ML architectures with traditional geophysical methodologies.
Labs/Teams: While specific lab affiliations are not detailed, their collaborations suggest involvement with UBC’s geophysical research groups and international partnerships with institutions like CNRS.
