Multi-Task Learning of Height and Semantics from Aerial Images - Université Paris Descartes (Paris 5) Access content directly
Journal Articles IEEE Geoscience and Remote Sensing Letters Year : 2019

Multi-Task Learning of Height and Semantics from Aerial Images

Apprentissage multi-tâche de l'élévation et de la sémantique à partir d'images aériennes

Abstract

Aerial or satellite imagery is a great source for land surface analysis, which might yield land use maps or elevation models. In this investigation, we present a neural network framework for learning semantics and local height together. We show how this joint multi-task learning benefits to each task on the large dataset of the 2018 Data Fusion Contest. Moreover, our framework also yields an uncertainty map which allows assessing the prediction of the model. Code is available at https://github.com/marcelampc/mtl_aerial_images .
Fichier principal
Vignette du fichier
DTIS19229.1580910909_postprint.pdf (6.63 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02386074 , version 1 (29-11-2019)
hal-02386074 , version 2 (10-02-2020)

Identifiers

Cite

Marcela Carvalho, Bertrand Le Saux, Pauline Trouvé-Peloux, Frédéric Champagnat, Andrés Almansa. Multi-Task Learning of Height and Semantics from Aerial Images. IEEE Geoscience and Remote Sensing Letters, 2019, ⟨10.1109/LGRS.2019.2947783⟩. ⟨hal-02386074v2⟩
103 View
76 Download

Altmetric

Share

Gmail Facebook X LinkedIn More