Accéder directement au contenu Accéder directement à la navigation
Communication dans un congrès

Adapted learning for polarization-based car detection

Abstract : Object detection in road scenes is an unavoidable task to develop autonomous vehicles and driving assistance systems. Deep neural networks have shown great performances using conventional imaging in ideal cases but they fail to properly detect objects in case of unstable scenes such as high reflections, occluded objects or small objects. Next to that, Polarized imaging, characterizing the light wave, can describe an object not only by its shape or color but also by its reflection properties. That feature is a reliable indicator of the physical nature of the object even under poor illumination or strong reflections. In this paper, we show how polarimetric images, combined with deep neural networks, contribute to enhance object detection in road scenes. Experimental results illustrate the effectiveness of the proposed framework at the end of this paper.
Liste complète des métadonnées

https://hal-univ-bourgogne.archives-ouvertes.fr/hal-02471097
Contributeur : Imvia - Université de Bourgogne <>
Soumis le : vendredi 7 février 2020 - 17:25:37
Dernière modification le : samedi 8 février 2020 - 01:32:14

Identifiants

Citation

Rachel Blin, Samia Ainouz, Stephane Canu, Fabrice Mériaudeau. Adapted learning for polarization-based car detection. 14th International Conference on Quality Control by Artificial Vision, May 2019, Mulhouse, France. pp.98, ⟨10.1117/12.2523388⟩. ⟨hal-02471097⟩

Partager

Métriques