Register now After registration you will be able to apply for this opportunity online.
Maximum flexion: a problem for markerless tracking?
This internship/research project/semester thesis elaborates why in various of our data collections markerless tracking in maximum flexion became very inaccurate.
We use markerless tracking to develop easy-to-implement motion analysis for applications ranging from physiotherapy at home to the media coverage of sports competitions. In previous data collections, in which we applied both markerless and marker-based method, the largest deviations between the two methods occurred in phases of maximum flexion. We now want to reproduce these deviations in order to better understand their cause so that we can outline how markerless tracking can be improved.
We use markerless tracking to develop easy-to-implement motion analysis for applications ranging from physiotherapy at home to the media coverage of sports competitions. In previous data collections, in which we applied both markerless and marker-based method, the largest deviations between the two methods occurred in phases of maximum flexion. We now want to reproduce these deviations in order to better understand their cause so that we can outline how markerless tracking can be improved.
- Becoming familiar with existing pose detection algorithms and relevant literature
- Understanding of existing data sets at our Lab
- Theoretically deriving sources of errors
- Reproducing the observed errors in climbing experiments in our Lab
- Demonstrating simple means to avoid errors (if possible)
- Becoming familiar with existing pose detection algorithms and relevant literature - Understanding of existing data sets at our Lab - Theoretically deriving sources of errors - Reproducing the observed errors in climbing experiments in our Lab - Demonstrating simple means to avoid errors (if possible)
Opportunity to delve into markerless tracking in specific applications and to plan and carry out your own experiments.
Opportunity to delve into markerless tracking in specific applications and to plan and carry out your own experiments.
Student of Health Sciences and Technology or Biomedical Engineering
Student of Health Sciences and Technology or Biomedical Engineering
Please apply via email to Peter Wolf, pwolf@ethz.ch.
Sensory-Motor Systems Lab, ETH Zurich
Please apply via email to Peter Wolf, pwolf@ethz.ch. Sensory-Motor Systems Lab, ETH Zurich
Ad_ErrorEllbowFlexion_2024.pdf
