PtyRANNOSAUR: Ptychography with Robust Artificial Neural Networks Optimized for Sub-Angstrom Accuracy and Ultrafast Reconstruction
Kieran Loehr, Rahim Raja, Xiaochuan Ding, Jeffrey Huang, Gillian Nolan, Sang hyun Bae, Bryan K. Clark, Pinshane Y. Huang · 2026
Read PtyRANNOSAUR: Ptychography with Robust Artificial Neural Networks Optimized for Sub-Angstrom Accuracy and Ultrafast Reconstruction by Kieran Loehr, Rahim Raja, Xiaochuan Ding, Jeffrey Huang, Gillian Nolan, Sang hyun Bae, Bryan K. Clark, Pinshane Y. Huang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-27587.txt.
Description
Electron ptychography images materials at atomic scale. It rebuilds images from diffraction patterns, but the standard process is slow. Kieran Loehr and seven colleagues present PtyRANNOSAUR in 2026. Their neural network code turns microscope data into phase images within seconds. It runs roughly ten to a hundred times faster than standard methods.
Convolutional autoencoders do the rebuilding. An autoencoder compresses data and then rebuilds it. Each model trains on a large database of crystal structures. Models are tuned to settings such as voltage, beam angle, defocus, and sample thickness. Users need no fine-tuning of their own. The code also handles partial coherence, multiple scattering, and scan position errors. Tests used both experimental and simulated data. The networks recovered atomic structures across many materials. Resolution reached finer than half an angstrom. That matches the best slow iterative reconstructions of the same data.
Catalog record
| CREATOR | Kieran Loehr, Rahim Raja, Xiaochuan Ding, Jeffrey Huang, Gillian Nolan, Sang hyun Bae, Bryan K. Clark, Pinshane Y. Huang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.27587