Internship: AI-based extraction of nanoparticle morphology from SEM images

Background

In magnetic tape recording, the magnetic layer consists of nanometer-sized magnetic particles whose shape, size, and aspect ratio strongly influence the medium's magnetic properties and the achievable storage density. Accurate characterization of these morphological properties is essential for micromagnetic simulations that predict media performance and support the development of next-generation tape media. Scanning Electron Microscopy (SEM) images of magnetic tape resolve individual particles, enabling direct morphological analysis. However, extracting statistically meaningful morphology distributions from thousands of particles remains a labor-intensive manual process.

This internship aims to develop an AI-based approach to automate nanoparticle morphology extraction from SEM images.

Project

This is a three-month internship at IBM Research – Zurich, starting in May or June 2027. The intern will:

• Review and compare state-of-the-art machine learning methods for nanoparticle morphology extraction.
• Develop an automated pipeline to extract shape, size, and aspect ratio distributions of magnetic recording nanoparticles from SEM image datasets.
• Benchmark the methods and pipeline against reference data, assessing both accuracy and processing efficiency.

Requirements

• Master's or PhD student in computer science, electrical engineering, physics, mathematics, or a related field.
• Good Python programming skills.
• Preferable: knowledge of/background in optimization, statistics, or machine learning.