Pose estimation and motion reconstruction for contact rich interactions

The goal of this project will be to develop new methodologies for accurately reconstructing human motion when there are significant amounts of contact with involved. Specifically, the project focuses on recovering poses for activities such as climbing, wrestling, and manipulation of objects, where human-human and human-environmental contact plays an important factor. The overall strategy will be to fuse multi-modal data captured from real world sensors (motion capture) with parametric and physics-based body models to improving accuracy and reconstruction.

Interested candidates should contact Prof. Andrews by sending the following items by email:
• Two-page CV ;
• Transcripts from a graduate program ;
• An example of a technical report or scientific article that they authored

Connaissances requises

- Master’s degree in computer science, software engineering or a related field
- Knowledge of video games, physics simulation, and 3D graphics
- Knowledge of supervised learning, reinforcement learning, and numerical optimization
- Experience with motion capture technology is helpful (cameras, IMUs, optical markers)
- Strong programming skills in C++ or Python
 

Programme d'études visé

Doctorat

Domaines de recherche

Technologies de l'information et des communications

Financement

Three (3) years funded with possibility to extend

Autres informations

Starting : 2021-08-29 (Affichage hiver - été 2021)

Personne à contacter

Sheldon Andrews | sheldon.andrews@etsmtl.ca