Clémentin Boittiaux, Vincent Lepetit

A Python Blender-based simulation framework designed to benchmark scene exploration, 3D reconstruction, and semantic segmentation tasks in both static and dynamic environments.

  arXiv project page   code   Blender file

Clémentin Boittiaux, Ricard Marxer, Claire Dune, Aurélien Arnaubec, Maxime Ferrera, Vincent Hugel

Use SfM/MVS dense reconstruction to recover the colors of underwater images.

  arXiv project page   code

Clémentin Boittiaux, Claire Dune, Maxime Ferrera, Aurélien Arnaubec, Ricard Marxer, Marjolaine Matabos, Loïc Van Audenhaege, Vincent Hugel

The dataset presents four dives on the Eiffel Tower hydrothermal vent over a five-year period. We provide images, navigation data and a unified SfM ground truth.

  arXiv   code   dataset

Clémentin Boittiaux, Ricard Marxer, Claire Dune, Aurélien Arnaubec, Vincent Hugel

Approximating the reprojection error through homographies addresses differentiability issues for camera pose regression in deep learning applications.

  arXiv   code video