Human body and limbs recognition and reconstruction

Masterarbeit

Themen-Schwerpunkt: Automation und Robotik, Maschinelles Lernen, Service-Robotik und mobile Manipulation
Studiengänge: Elektrotechnik, Informatik, Mechatronik

Umfeld

Human body recognition is a critical skill in human-robot collaboration, as robots are expected to be in closer contact with the user and higher reliability is required.

However, the current state of art for human body recognition requires the full body picture, and it does not allow to detect only body parts or to reconstruct the full body from limb observation. As it is still an open question, we are searching for a student interested in researching feasible solutions for human body segmentation, body limbs recognition and human body reconstruction.

Aufgaben

  • Develop and implement new algorithms for human body segmentation and reconstruction
  • Recognice and predict with high accuracy human limbs, independently of clothing, light conditions and another occlusion sources
  • Work closely with another team members in the project

Wir bieten

  • The opportunity to gain experience in a technical-scientific environment
  • Insights into the area of human-robot interaction
  • An interdisciplinary working environment with partners from science, industry and society
  • A pleasant working atmosphere
  • Constructive teamwork

Wir erwarten

  • Good knowledge of Python or C++
  • Experience with Linux
  • Experience with Computer Vision
  • Experience with Machine Learning is a plus
  • Self-reliant thinking and working
  • Very good knowledge of English
  • Motivation and commitment

Weitere Informationen

  • Start: as soon as possible
  • Supervising institute at KIT: Institute fuer Anthropomatik (HIS) | Prof. Dr.-Ing. Rüdiger Dillmann
  • Contact: Dr. Genessis Perez, perez@dont-want-spam.fzi.de