About the job
In this Master thesis, you will contribute to the development of advanced anomaly detection methods for industrial multi-channel vision systems for quality assessment in series production. Our goal is to design a generalizable deep-learning model that leverages vision features to reliably detect defects across varying products and manufacturing scenarios.
Responsibilities
Explore cutting-edge approaches in anomaly detection and multi-channel image processing to develop a novel, generalizable model for quality assessment in industrial production.
Implement, train, and evaluate your approach using real-world production data, ensuring its robustness and suitability for industrial applications.
Work closely with experts from research and development as well as production, regularly presenting your results and thus fostering continuous innovation in computer vision for manufacturing.
Qualifications
Minimum
Education: advanced studies in the field of Computer Science, Machine Learning, Artificial Intelligence or comparable
Experience and Knowledge: excellent machine-learning fundamentals and very good programming skills in Python and in one deep learning framework, e.g. PyTorch or JAX
Personality and Working Practice: you are able to approach complex tasks in a structured and analytical manner, you have a high motivation to learn and work independently on challenging topics, always communicating your results clearly and understandably
Work Routine: we offer you the possibility to work remotely part of the time
Languages: very good in English
Preferred
No preferred qualifications listed.