Abdullah Haroon Rasheed

Primarily working on applying machine learning techniques to data generated by physics based mechanistic simulators of slender structures in order to solve inverse problems. The work focuses on the following:


  • Inverse measurement problem for visual measurement of the static friction coefficient from cloth contact.
  • Inverse design problem of measuring the deformed curvature of a suspended curly rod, in order to get the natural shape of the rod.
  • Validation of simulators against physical experiments to ensure that data generated from them is physically plausible and captures the required phenomena accurately.
  • Training of machine learning models on simulated data, and testing on real data captured through a physical experiment in a controlled setting.

The work is a collaboration between ELAN and Morpheo teams at INRIA Grenoble. 

Under the supervision of Florence Bertails-Descoubes, Stefanie Wuhrer and Jean-Sebastien Franco.





   This research is funded by ERC grant GEM (StG-2014-639139)