https://github.com/ArdaGen/4D-STEM-neural-diffraction-pattern-recognition-tempo4d
High-throughput analysis of Bragg discs from 4D-STEM datasets using ML object detection.
https://github.com/ArdaGen/4D-STEM-neural-diffraction-pattern-recognition-tempo4d
Science Score: 36.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (13.0%) to scientific vocabulary
Keywords
Repository
High-throughput analysis of Bragg discs from 4D-STEM datasets using ML object detection.
Basic Info
Statistics
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
- Releases: 1
Topics
Metadata Files
README.md
tempo4D
Transmission Electron Microscope Pattern Observation
This repository provides an automated pipeline for analyzing 4D-STEM datasets using YOLOv8n.
The workflow enables end-to-end processing of large-scale 4D-STEM datasets
for phase identification, orientation mapping (coming soon!), and strain analysis.
🧬 Phase Mapping
Phase mapping of complex phase-transformed Ti-50Nb alloy using object detection-based pattern recognition.
🧪 Strain Mapping
Strain mapping of Si/SiGe multilayers demonstrating coherent lattice mismatch analysis.
Supported file formats:
- Thermo Fisher Scientific:
.emi,.xml(EMPAD) - GATAN:
.dm3,.dm4 - Dectris:
.h5 - NanoMegas:
.blo - Direct Electron:
.de5 - Standard:
.h5,.hdf5
🛠️ Installation
Python ≥ 3.9 is required.
We recommend creating a new virtual environment:
bash
conda create -n tempo4d python=3.9
conda activate tempo4d
⚡ Install PyTorch (Recommended First)
If you have a CUDA-capable GPU, install a CUDA-compatible version of PyTorch before installing tempo4d.
👉 Install PyTorch
📦 Install tempo4d
pip install tempo4d
This will install all required dependencies, including:
- PyQt5
- pyqtgraph
- OpenCV
- matplotlib
- Ultralytics (for YOLOv8)
- rosettasciio[all] (for TEM file support)
Demo
Please also see the tempo4d_demo.ipynb notebook in the demo folder.
Download example data from GATAN
Cite
``` @misc{genc2025neuralobjectdetection4d, title={Neural Object Detection for 4D STEM: High-Throughput Sub-Pixel Electron Diffraction Pattern Recognition}, author={Arda Genc and Ravit Silverstein}, year={2025}, eprint={2506.04477}, archivePrefix={arXiv}, primaryClass={cond-mat.mtrl-sci}, url={https://arxiv.org/abs/2506.04477}, }
```
Owner
- Login: ArdaGen
- Kind: user
- Repositories: 1
- Profile: https://github.com/ArdaGen