Master's Thesis: Development of an Intelligent Superclassifier for the Optimization of AI Models (m/w/d)

ATLAS ELEKTRONIK is a company of thyssenkrupp AG and stands for maritime safety. For decades, as a leading system house for maritime electronics marines around the world, we have been helping to make the lake safer. We offer a wide range of services to our customers including Sonars and Sensors, Guide Systems for Unmanned Underwater Vehicles, Submarines and Overwater Ships, Mine Defense Systems and Radio and Communication Systems.

Analysis of a preparatory work on AI models in the basic classification to identify suitable AI models for integration. Superclassifier development: Conception and implementation of a hybrid classifier using Few-Shot Learning methods to refine and generalize the results of a pretrained AI model. Experimental validation and evaluation of the developed superclassifier. These tasks are documented as part of your Master's thesis.

You can study computer science, data science, machine learning or comparable. You have very good Python programming skills and have already gained first experience in deep learning. You can work yourself into new topics. You have a high willingness to introduce yourself to the A-LAB team and to actively shape the student lab in an interdisciplinary environment. You're a creative mind and you're drawing yourself through a communicative way that you bring in various specialist discussions. You have very good knowledge of German and good English in word and writing.

Through the A-LAB, ATLAS ELEKTRONIK offers doctoral students and students the opportunity to work in an agile environment with an interdisciplinary team on scientific topics with high application relevance. As part of your doctorate, you will get the opportunity to work together with students on your scientific topic. The doctoral program of the ATLAS ELEKTRONIK supports you both technically, professionally and with expertise in the doctoral procedure with a respective university to be determined.

We value diversity

We value diversity and therefore welcome all applications – regardless of gender, nationality, ethnic and social origin, religion/worldview, disability, age as well as sexual orientation and identity. JOBV1_EN