Machine Learning Models for Atomic Scale Imaging
-
Job Type:
Bachelor/Master Thesis
-
Contact Person:
Dr. Philip Willke (Contact us for more information)
Machine Learning Models for Atomic Scale Imaging
Bachelor/Master Thesis Project
Do you..
- enjoy programming and writing your own code?
- want hands-on experience with machine learning and neural networks?
- yant to work with experiments on Quantum Nanoscience at the atomic scale?
- plan to write your bachelor’s / master’s thesis soon?
Project overview
Our group images and manipulates atoms and molecules using a scanning tunneling microscope (STM). In this thesis, you will help automate and optimize STM measurements with machine learning - using experimental data to detect nanoscale objects and to run measurement routines with minimal human intervention.
What you will do
- Acquire STM datasets that resolve individual atoms and molecules.
- Learn to operate an STM and work with ultrahigh vacuum (UHV) conditions / low temperatures.
- Build end-to-end ML workflows in Python (e.g., PyTorch): data pipelines, model development, training, evaluation, and performance improvement.
What you will learn
- Practical STM operation, electronics, UHV basics, and advanced data acquisition.
- Experimental data analysis and evaluation.
- Modern ML tooling for scientific data
Tools & skills
- Python; ML libraries such as PyTorch; standard scientific Python stack.
- Version control and clean, reproducible code practices.
Contact
Write to Prof. Dr. Philip Willke (PHI, KIT) philip.willke∂kit.edu if interested.
More information: https://www.willkelab.com
