We are thrilled to announce the successful conclusion of the first-ever Machine Learning for Electron and Scanning Probe Microscopy Hackathon, hosted at the University of Tennessee, Knoxville. This event brought together a vibrant and diverse group of participants to tackle cutting-edge challenges in microscopy through machine learning.
Participation:
The hackathon saw 20 submissions from 70-80 participants across the globe. Many participants worked on multiple projects, showcasing their dedication and versatility.
Global Reach:
The event had an international flavor with participants from renowned institutions such as Cornell, LBL, Purdue, Clemson, PSU, Hong Kong U, UniGe, DTU, U Gießen, U Chile, Uni Mainz, ICMAB, SolidPower Battery, Argonne, USC, UT Knoxville, and many more. Over two-thirds of the teams comprised members from multiple institutions, highlighting the collaborative spirit of this hackathon.
Projects:
Participants submitted innovative projects, ranging from reward-driven phase mapping to automated nanoparticle detection, AI-powered thermal mapping, and unsupervised classification of ferroelectric domains.
Winners
A big round of applause for the winners of this inaugural hackathon:
1st Place:
GANder: Ferroelastic–Ferroelectric Domains Observed by Image-to-Image Translation
Team Members: Ralph Bulanadi (University of Geneva, team lead), Kieran J Pang (Justus-Liebig-Universität Gießen), Michelle Wang (Technical University of Denmark)
2nd Place:
AutoScript Copilot
Team Members: Xiangyu Yin (Argonne National Laboratory), Yi Jiang (Argonne National Laboratory), Yu-Tsun Shao (University of Southern California), Benjamin Fein-Ashley (University of Southern California)
3rd Place:
Structure Discovery through Image-to-Graph Machine Learning Model
Team Members: Lauri Kurki (Aalto University, team lead), Harshit Sethi (Aalto University), Jie Huang (Aalto University)
Student Council Award:
Reward-Based Segmentation: Phase Mapping of 2D Polycrystalline Pd-Se Phases
Team Members: Kamyar Barakati (University of Tennessee, team lead), Aditya Raghavan (University of Tennessee)
Honorable Mentions:
- Microscopy LLM: Adib Bazgir (University of Missouri-Columbia), Rama Chandra Praneeth Madugula (New York University), Yuwen Zhang (University of Missouri-Columbia)
- Automating AFM through Model-Driven Image Segmentation and Classification: Sam Welborn (NERSC/NCEM, team lead), Mikolaj Jakowski (UTK), Shawn Patrick (UTK), Alex Pattison (NCEM), Panos Manganaris, Sirisha Madugula (ORNL)
- Unmasking Biomacromolecular Conformational Dynamics from 2D Analysis of Subdomains Dynamic Modes and Molecular Kinetics: Ian Addison-Smith, Willy Menacho, and Horacio V. Guzman (Institut de Ciència de Materials de Barcelona (ICMAB-CSIC), Barcelona, Spain)
Special Thanks
We extend our heartfelt gratitude to the organizers: Gerd Duscher, Rama Vasudevan, Steven R. Spurgeon, Colin Ophus, Utkarsh Pratiush, Austin Houston, Yongtao Liu, Yu Liu, and Maxim Ziatdinov.
Additionally, we thank our sponsors: Office of Naval Research, Thermo Fisher Scientific, AI Tennessee, Vasileios Maroulas, and our partnering organizations.
Next Steps
All project details and codes are now available here. Stay tuned for a detailed sharing of the projects and results!
Let’s continue to push the boundaries of microscopy and machine learning!