Machine learning methods for automated analysis of four-dimensional diffraction datasets (4D-DD).

Applications:

1. Precision crystal orientation mapping
2. Strain mapping

Features:

1. A new approach to map crystal orientation and strain in TEM samples using trained artificial neural networks.
2. Use simulated dynamical electron diffraction intensity as the training dataset.
3. Work on large 4D-DD datasets with much faster processing speed compared with conventional methods.
4. High angular resolution at 0.009˚ for precision orientation mapping and strain sensitivity at 0.04% are achieved.

Example data: Raw data of simulated and experimental diffraction datasets exceed 10 GB. Only preprocessed data with greatly reduced size are included here as an example. Please contact Prof. Jian-Min Zuo (jianzuo (at) illinois (dot) edu) for raw data.

If you find this software package is useful, please cite: Renliang Yuan, Jiong Zhang, Lingfeng He, Jian-Min Zuo, Training artificial neural networks for precision orientation and strain mapping using 4D electron diffraction datasets, Ultramicroscopy (2021).