PhD and Postdoc positions project HuMoCar (Strasbourg)

This is an announcement for a PhD and a postdoctoral position funded by the bi-national and tri-institutional project "HuMoCar: Human Model for Real-World Human Cognition by Care Robots".

We seek candidates for a PhD and a postdoctoral researcher position in Computer Science, with backgrounds/expertise in one or several of the following areas:

- Computer Vision for Robotics

- Deep Learning for Computer Graphics

- Computational Mechanics

Hosting institute:

ICube Laboratory (Le laboratoire des sciences de l'ingénieur, de l'informatique et de l'imagerie : The Engineering science, computer science and imaging laboratory) at the University of Strasbourg is a leading research center in Computer Science, with more than 300 permanent researchers, with the recently opened AI graduate school supported by the French government.

Application process:

- Deadline for applications is on the 30th of January 2022, for a starting date on the 1st of March 2022 (can be negotiated).

- Candidates should submit their applications on https://emploi.cnrs.fr/.

- The gross salary of the PhD candidate is approximately 2135€/month, for a duration of 3 years.

- The salary of the postdoc researcher will depend on the successful candidate's prior research experience with a minimum gross salary of 2675 €/month.

Project summary

We aim to push the current limits of robot vision in human cognition by care-robots in the in-house situation. Our specific goal is to make the performance of the vision-intelligence robust to large variations (in body shapes, motions,..) to occlusion (cloth, furniture, wall,..), and capable of understanding the interaction by developing a photo-realistic, physics-aware 4D human model.

* Main tasks

Research and development in one or more of the following topics, with possible supervision of a PhD student:

− Photo-realistic human modeling: Extension of the geometric human body model with color and illumination.

− Action-conditioned motion prediction/generation model via deep-learning over annotated motion datasets.

− End-to-end 4D human model reconstruction from 2D/3D video input via optimal model fitting.

− Physics-aware human-object interaction: Integration of 4D human model with a PBM and development of interaction motion controller DNNs.

More informations can be found on the following webpage:

http://igg.unistra.fr/People/seo/StudentsJoboffer_files/Postdoc_2022.pdf

http://igg.unistra.fr/People/seo/StudentsJoboffer_files/PhD_2022.pdf