Hackathon 2 - Data curation and unsupervised analysis of the diffraction data - Houston

Hackathon 2 - Data curation and unsupervised analysis of the diffraction data - Houston

by Machine Learning in the Nanoworld

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQMMachine learning
🎙 Machine Learning in the Nanoworld 👥 1K 📅 July 18, 2026 ⏱ 17 min 👁 24 📄 tutorial 🧭 2026-08-16
Available in: English (current) Français

Keywords

4D-STEMdigital twindiffractionlattice parameterunsupervised analysis

Summary

This video is a recording of a hackathon session from a workshop on machine learning in the nanoworld. The presenter guides participants through a Jupyter notebook that simulates 4D-STEM diffraction pattern acquisition using a digital twin environment based on asynchroscopy. The session covers setting up the environment, acquiring overview images, placing the electron probe, and obtaining diffraction patterns. Participants are provided with tools for particle segmentation and lattice parameter extraction. The hackathon includes four exercises of increasing difficulty: comparing center-to-edge lattice profiles, reconstructing lattice parameter maps using sparse measurements, defining smarter sampling grids, and mapping the entire field of view. The presenter reveals that oval particles have a larger lattice parameter than circular ones, with variations across the particles. The session concludes with a discussion of patterns observed and an invitation to run solutions on a real microscope.

138 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable hands-on experience in using digital twins for scientific data analysis, demonstrating a practical workflow for 4D-STEM data. The argumentation is based on direct demonstration and practical problem-solving, with clear explanations of the tools and methods. The presenter encourages participants to explore and discover patterns, fostering an interactive learning environment. The value lies in the transferable skills and the open-source nature of the tools used.

Scientific Rigor, Source Quality, Title Accuracy

The video is scientifically rigorous in its methodology, using established simulation tools and open-source software. However, it does not cite specific scientific sources or publications, relying instead on the workshop context and the tools themselves. The title accurately reflects the content, which is a hackathon session focused on data curation and unsupervised analysis. No comments were provided for analysis.

142 words

Title / Content Match

The title accurately describes the content: a hackathon session on data curation and unsupervised analysis of diffraction data.

Quality & Reliability

7/10

The video is a hands-on tutorial from a workshop, demonstrating practical use of open-source tools (asynchroscopy, digital twin) for 4D-STEM data analysis. It is technically sound and reproducible, but lacks formal citations and peer review.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

  • asynchroscopy GitHub repository — The digital twin and data acquisition tools are based on asynchroscopy.
  • ABTEM — Used for simulating diffraction patterns.
  • Atomic Simulation Environment (ASE) — Used for atomic simulations.

Concurring Sources

  • asynchroscopy GitHub repository — The digital twin and data acquisition tools are based on asynchroscopy.

Contribution & Novelties

The video demonstrates a novel approach to teaching and performing 4D-STEM data analysis using a digital twin, which allows for hands-on experimentation without the need for a physical microscope. This approach is highly scalable and accessible, enabling researchers to develop and test analysis algorithms in a simulated environment before applying them to real data. The hackathon format encourages active learning and problem-solving, with a focus on unsupervised analysis and pattern discovery.

Pour aller plus loin :

  • 4D-STEM — Overview of 4D-STEM technique.
  • Digital twin — Concept of digital twins in engineering.
  • asynchroscopy documentation — Official documentation for the asynchroscopy library.
  • ABTEM — Simulation tool for electron microscopy.
  • Atomic Simulation Environment — Tool for atomic-scale simulations.

115 words

Radar Profile

The radar profile shows high scores in technical level and information quality, reflecting the advanced nature of the content and its practical value. The lower score in information quantity is due to the short duration and focused scope. Overall, the video is a solid technical resource for those interested in 4D-STEM data analysis.

Reliability 7/10