Master Thesis: Machine Learning Consumer Radar (f/m/div)*

  • Infineon Technologies
  • München, Germany
  • 14/09/2020
Full time Data Science Data Analytics Big Data Data Management Statistics

Job Description

At a glance

You are right about to finish you Master studies in Computer Science or a similar subject? Now you are on the lookout for a practical Master thesis topic that puts a cherry on your degree? Read on, we have an interesting topic to offer! An essential factor for smart home or smart office applications is to know whether people are located in a room. Although computer vision (CV) systems yield a high accuracy, radar based solutions are preferred due to privacy issues. An especially difficult scenario for radar-based people sensing is when there are multiple vibrating sources such as fans, ventilators and so on. In this case, the signal has to be analyzed more accurately which comes with higher computational costs. In order to keep the computational complexity small, a way of solving this problem is to detect a target region and analyze only the specific region in detail, such as by region proposal networks (RPN). As a result, the main goal of the master thesis is to develop a RPN based People Sensing solution using only raw radar data, that is able to propose target regions, which are then analyzed with high accuracy in order to yield an accurate people detection rejecting other motions within a room. Sounds interesting? Apply now!

Job description

The goal of this thesis is to develop a machine learning based People Sensing solution using radar data that is able to classify if there are people in a room. To guarantee high accuracy, vibrating sources such as fans, ventilators and so on need to be rejected. This includes:

  • Literature research;
  • Data recording and label generation in Data Acquisition Setup;
  • Implementation of a prove of concept using state-of-the-art architectures;
  • Evaluation of accuracy and computational complexity.

Your Profile

You are best equipped for this task if you:

  • Will soon complete your Master studies in Computer Science or a similar subject;
  • Already gained experience with digital signal processing;
  • Bring along a good understanding of basic machine learning and deep learning concepts;
  • Are familiar with machine learning frameworks and tools like TensorFlow, PyTorch;
  • Have good Python programming skills.

Please attach the following documents to your application:

  • CV in English;
  • Certificate of enrollment at university;
  • Latest grades transcript;
  • High school report.


What we offer you at Campeon

In our headquarter in Neubiberg near Munich, more than 3.300 employees are working in research & development, several central functions, IT and many more.