Berta Bescos — accepted AI-conference papers

2022

Don't Share My Face: Privacy Preserving Inpainting for Visual Localization

Saad Himmi, Oguzhan Ilter, François Pailleau, Roland Siegwart, Berta Bescos, Cesar Cadena

IROS 2022poster

Visual localization is an important task for many robotic and augmented reality applications. As localizing within large scale maps can be memory and computationally de-manding, cloud-based localization services are appealing for developers. However, such services raise important privacy concerns fo…

Cited by 4 Source ↗Scholar ↗

2019

Empty Cities: Image Inpainting for a Dynamic-Object-Invariant Space

Berta Bescos, José Neira, Roland Siegwart, Cesar Cadena

ICRA 2019poster

In this paper we present an end-to-end deep learning framework to turn images that show dynamic content, such as vehicles or pedestrians, into realistic static frames. This objective encounters two main challenges: detecting all the dynamic objects, and inpainting the static occluded background with…

Cited by 38 Source ↗code ↗Scholar ↗