STM-Net: Reconstructing Molecular Orbitals from STM Images
Jan 1, 2024
·
1 min read
In a nutshell: Are orbitals observable?
This project reconstructs pristine molecular-orbital information from high-resolution scanning tunneling microscopy images. STM images are simulated at the density-functional-theory level using Bardeen’s approximation, and STM-Net separates s-wave molecular-orbital signals from p-wave contributions introduced by CO-functionalized tips.
The approach connects Tersoff-Hamann theory and Chen’s derivative rule with physics-informed image restoration. Experiments included DAE, U-Net, and Mask2Former-based models.

Authors
AI for Science · Computational Chemistry · Computer Vision
Yu Zhu works at the intersection of AI for Science, computational chemistry,
and computer vision. His research spans machine learning force fields,
physics-driven reconstruction of molecular orbitals from scanning tunneling
microscopy images, open-set fine-grained recognition, and materials modeling.