Die u:cris Detailansicht:

Deep reinforcement learning for data-driven adaptive scanning in ptychography

Autor(en)
Marcel Schloz, Johannes Müller, Thomas C. Pekin, Wouter Van den Broek, Jacob Madsen, Toma Susi, Christoph T. Koch
Abstrakt

We present a method that lowers the dose required for an electron ptychographic reconstruction by adaptively scanning the specimen, thereby providing the required spatial information redundancy in the regions of highest importance. The proposed method is built upon a deep learning model that is trained by reinforcement learning, using prior knowledge of the specimen structure from training data sets. We show that using adaptive scanning for electron ptychography outperforms alternative low-dose ptychography experiments in terms of reconstruction resolution and quality.

Organisation(en)
Physik Nanostrukturierter Materialien
Externe Organisation(en)
Humboldt-Universität zu Berlin
Journal
Scientific Reports
Band
13
Anzahl der Seiten
10
ISSN
2045-2322
DOI
https://doi.org/10.48550/arXiv.2203.15413
Publikationsdatum
05-2023
Peer-reviewed
Ja
ÖFOS 2012
103018 Materialphysik, 103042 Elektronenmikroskopie, 102019 Machine Learning
ASJC Scopus Sachgebiete
General
Link zum Portal
https://ucrisportal.univie.ac.at/de/publications/ffe03a04-074d-4b55-a172-4ef95a6b1e39