Self-supervised learning for denoising quasiparticle interference data

作     者：Ilse S. Kuijf Willem O. Tromp Tjerk Benschop Niño Philip Ramones Miguel Antonio Sulangi Evert P. L. van Nieuwenburg Milan P. Allan 

作者机构：Leiden Institute of Physics Leiden University 2300 RA Leiden The Netherlands 〈aQa〉 at Lorentz Institute and Leiden Institute of Advanced Computer Science Leiden University P.O. Box 9506 2300 RA Leiden The Netherlands National Institute of Physics University of the Philippines Diliman Quezon City 1101 Philippines Faculty of Physics Ludwig-Maximilians-University Munich 80799 Munich Germany Center for Nano Science (CeNS) Ludwig-Maximilians-University Munich Munich 80799 Germany Munich Center for Quantum Science and Technology (MCQST) Ludwig-Maximilians-University Munich Munich 80799 Germany 

出 版 物：《Physical Review B》 (Phys. Rev. B)

年 卷 期：2025年第111卷第3期

页      面：035136-035136页

核心收录：

基　　金：Dutch National Growth Fund National Gaucher Foundation, NGF Quantum Delta NL 

主　　题：Self supervised learning 

摘      要：Tunneling spectroscopy is an important tool for the study of both real- and momentum-space electronic structure of correlated electron systems. However, such measurements often yield noisy data. Machine learning provides techniques to reduce the noise in postprocessing, but traditionally requires noiseless examples which are unavailable for scientific experiments. In this work we adapt the unsupervised Noise2Noise and self-supervised Noise2Self algorithms, which allow for denoising without clean examples, to denoise quasiparticle interference data. We first apply the techniques on simulated data, and demonstrate that we are able to reduce the noise while preserving finer details, all while outperforming more traditional denoising techniques. We then apply the Noise2Self technique to experimental data from an overdoped cuprate [(Pb,Bi)2Sr2CuO6+δ] sample. Denoising enhances the clarity of quasiparticle interference patterns, and helps to obtain a precise extraction of electronic structure parameters. Self-supervised denoising is a promising tool for denoising quasiparticle interference data, facilitating deeper insights into the physics of complex materials.