You are currently viewing Thesis Topic – Optimization of a neural network for motion tracking
Machine learning , artificial intelligence, ai, deep learning blockchain neural network concept. Brain made with shining wireframe above multiple blockchain cpu on circuit board 3d render.

Thesis Topic – Optimization of a neural network for motion tracking

Part Time
Regensburg
Posted 1 year ago

For our Regensburg location, we are offering a thesis at the earliest possible start date. As part of a thesis, an existing neural network for recognizing human poses is to be optimized. The improvement approaches are to be evaluated.

 

Detailed Description

  • Enrich key points or movement points
  • Apply methods of transfer learning
  • Improvement of calibration
  • Consider additional 2D cameras
  • Volumetric triangulation
  • Consider 3D point cloud

The use case is similar to Open Pose (https://github.com/CMU-Perceptual-Computing-Lab/openpose).

 

Your Profile

  • Student studying Computer Science, Electrical Engineering or similar subjects
  • Good knowledge of C++, Java or C#
  • Knowledge in machine learning and image processing
  • ROS, OpenCV and Linux knowhow is beneficial
  • Good knowledge of the German and the English language
  • Drive, flexibility, creativity, commitment, ability to work in a team, to communicate and organizational talent

Please take care to upload your current certificate of enrollment and also your current performance record, because they are obligatory for processing your application.

Applications from severely handicapped people are welcome.

 

Our Offer

  • Extensive design options and independent work in an international working environment characterized by flat hierarchies
  • Good working atmosphere, characterized by professional and personal appreciation
  • Dynamic team with great team spirit
  • Integration Program in a professional, young & dynamic team
  • Stability of a global operating company

Interested?

We Generate Technical Solutions
Pragmatic. Versatile. Responsive

Job Features

Job CategoryFor Students

Apply Online

A valid email address is required.
A valid phone number is required.

Zip Code / City