Etude des caméras light field à rolling shutter / Study of rolling shutter light field cameras
15 Novembre 2023
Catégorie : Stagiaire
Contact :Hermes McGRIFF, Renato Martins
Email: email@example.com Email: firstname.lastname@example.org
Place : ICB Laboratory (Dijon), Institut Marey - Maison de la Métallurgie
Initial training at Bac+5 in computer science or robotics, with prior knowledge in computer vision.
Taste for programming (especially in python) and experimentation.
Knowledge of using Blender software is a plus.
Knowledge of machine learning is a plus.
Good level of written and spoken English and French.
Creativity, imagination and curiosity, interest in the world of academic research. Reliability.
Particular attention will be given to issues of gender and social equity.
Application: Send CV + cover letter to email@example.com and Renato.firstname.lastname@example.org
Duration: 16 to 24 weeks from February-March 2024
Context: Light field cameras are cameras that capture the orientation of the rays of light entering its optical system . This makes it possible to recover the 3D geometry of the observed scene from a single image. Our study currently focuses on the behavior of this type of sensor when the scene is moving during the acquisition of this light field, considering that the camera is equipped with a rolling shutter sensor, and that, consequently, all the pixels are not exposed at the same time .
This internship is part of a project carried out between the ICB Laboratory and the Femto-st Institute (Besançon).
Mission: The recruited intern will be tasked with exploring one of the issues encountered by the interaction between light field and rolling shutter. The precise subject will be set up with the candidate before the start of the internship according to their preferences and skills profile. The potential missions are quite diverse in the field of computer vision, like for example:
- sizing the optical assembly of an experimental light field camera
- the application of machine learning methods for estimating depth maps adapted to the problem
- the application to robotics (probably in simulation)
 Edward H Adelson and John YA Wang. Single lens stereo with a plenoptic camera. PAMI, 1992.
 Omar Ait-Aider, Nicolas Andreff, Jean Marc Lavest, and Philippe Martinet. Simultaneous object pose and velocity computation using a single view from a rolling shutter camera. ECCV, 2006.
 Hermes Mcgriff, Renato Martins, Nicolas Andreff, Cédric Demonceaux. Joint 3D Shape and Motion Estimation from Rolling Shutter Light-Field Images, WACV, 2024