SemI2I: Semantically Consistent Image-to-Image Translation for Domain Adaptation of Remote Sensing Data - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year :

SemI2I: Semantically Consistent Image-to-Image Translation for Domain Adaptation of Remote Sensing Data

Abstract

Although convolutional neural networks have been proven to be an effective tool to generate high quality maps from remote sensing images, their performance significantly deteriorates when there exists a large domain shift between training and test data. To address this issue, we propose a new data augmentation approach that transfers the style of test data to training data using generative adversarial networks. Our semantic segmentation framework consists in first training a U-net from the real training data and then fine-tuning it on the test stylized fake training data generated by the proposed approach. Our experimental results prove that our framework outperforms the existing domain adaptation methods.
Fichier principal
Vignette du fichier
SemI2I.pdf (2.45 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02487142 , version 1 (21-02-2020)

Identifiers

  • HAL Id : hal-02487142 , version 1

Cite

Onur Tasar, S L Happy, Yuliya Tarabalka, Pierre Alliez. SemI2I: Semantically Consistent Image-to-Image Translation for Domain Adaptation of Remote Sensing Data. IGARSS 2020 - IEEE International Geoscience and Remote Sensing Symposium, Sep 2020, Waikoloa, Hawaii, United States. ⟨hal-02487142⟩
97 View
165 Download

Share

Gmail Facebook Twitter LinkedIn More