Homography-based loss function for camera pose regression

Some recent visual-based relocalization algorithms rely on deep-learning methods to perform camera pose regression from image data. This paper focuses on the loss functions that embed the error between two poses to perform deep-learning based camera pose regression. Existing loss functions are either difficult-to-tune multi-objective functions or present unstable reprojection errors that rely on ground-truth 3D scene points and require a two-step training. To deal with these issues, we introduce a novel loss function which is based on a multiplane homography integration. This new function does not require prior initialization and only depends on physically interpretable hyperparameters. Furthermore, the experiments carried out on well established relocalization datasets show that it minimizes best the mean square reprojection error during training when compared with existing loss functions.


Localization, deep learning for visual perception

Full Text

Publisher's official version
81 Mo
arXiv:2205.01937v1 - Preprint
8974 Ko
How to cite
Boittiaux Clementin, Marxer Ricard, Dune Claire, Arnaubec Aurelien, Hugel Vincent (2022). Homography-based loss function for camera pose regression. Ieee Robotics And Automation Letters. 7 (3). 6242-6249. https://doi.org/10.1109/LRA.2022.3168329, https://archimer.ifremer.fr/doc/00766/87810/

Copy this text