Large-scale semantic classification: outcome of the first year of Inria aerial image labeling benchmark

Bohao Huang (1) , Kangkang Lu (2) , Nicolas Audebert (3, 4) , Andrew Khalel (5) , Yuliya Tarabalka (6, 7) , Jordan Malof (1) , Alexandre Boulch (4) , Bertrand Le Saux (4) , Leslie Collins (1) , Kyle Bradbury (1) , Sébastien Lefèvre (8) , Motaz El-Saban (5)

1 Duke University [Durham]
2 NUS - National University of Singapore
3 OBELIX - Environment observation with complex imagery
4 DTIS, ONERA, Université Paris Saclay (COmUE) [Palaiseau]
5 Raisa Energy
6 TITANE - Geometric Modeling of 3D Environments
7 COMUE UCA - COMUE Université Côte d'Azur (2015-2019)
8 LIFAT - Laboratoire d'Informatique Fondamentale et Appliquée de Tours

Over the recent years, there has been an increasing interest in large-scale classification of remote sensing images. In this context, the Inria Aerial Image Labeling Benchmark has been released online in December 2016. In this paper, we discuss the outcomes of the first year of the benchmark contest, which consisted in dense labeling of aerial images into building / not building classes, covering areas of five cities not present in the training set. We present four methods with the highest numerical accuracies, all four being convolutional neural network approaches. It is remarkable that three of these methods use the U-net architecture, which has thus proven to become a new standard in image dense labeling.

Mots clés: classification benchmark; aerial images; U-net; deep learning; convolutional neural networks

Domaines: Traitement du signal et de l'image [eess.SP]

Bohao Huang, Kangkang Lu, Nicolas Audebert, Andrew Khalel, Yuliya Tarabalka, et al.. Large-scale semantic classification: outcome of the first year of Inria aerial image labeling benchmark. IGARSS 2018 - IEEE International Geoscience and Remote Sensing Symposium, Jul 2018, Valencia, Spain. pp.1-4, ⟨10.1109/IGARSS.2018.8518525⟩. ⟨hal-01767807⟩