TY - JOUR

T1 - An unsupervised domain adaptation approach for change detection and its application to deforestation mapping in tropical biomes

AU - Soto Vega, Pedro Juan

AU - Costa, Gilson Alexandre Ostwald Pedro da

AU - Feitosa, Raul Queiroz

AU - Ortega Adarme, Mabel Ximena

AU - Almeida, Claudio Aparecido de

AU - Heipke, Christian

AU - Rottensteiner, Franz

N1 - Funding Information: This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Deutsche Akademische Austauschdienst (DAAD), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), and NVIDIA Corporation.

PY - 2021/11

Y1 - 2021/11

N2 - Changes in environmental conditions, geographical variability and different sensor properties typically make it almost impossible to employ previously trained classifiers for new data without a significant drop in classification accuracy. Domain adaptation (DA) techniques been proven useful to alleviate that problem. In particular, appearance adaptation techniques may be used to adapt images from a specific dataset in such a way that the generated images have a style that is similar to the images from another dataset. Such techniques are, however, prone to creating artifacts that hinder proper classification of the adapted images. In this work we propose an unsupervised DA approach for change detection tasks, which is based on a particular appearance adaptation method: the Cycle-Consistent Generative Adversarial Network (CycleGAN). Specifically, we extend that method by introducing additional constraints in the training phase of the model components, which make it preserve the semantic structure and class transitions in the adapted images. We evaluate the proposed approach on a deforestation detection application, considering different sites in the Amazon rain-forest and in the Brazilian Cerrado (savanna) using Landsat-8 images. In the experiments, each site corresponds to a domain, and the accuracy of a classifier trained with images and references from one (source) domain is measured in the classification of another (target) domain. The results show that the proposed approach is successful in producing artifact-free adapted images, which can be satisfactory classified by the pre-trained source classifiers. On average, the accuracies achieved in the classification of the adapted images outperformed the baselines (when no adaptation was made) by 7.1% in terms of mean average precision, and 9.1% in terms of F1-Score. To the best of our knowledge, the proposed method is the first unsupervised domain adaptation approach devised for change detection.

AB - Changes in environmental conditions, geographical variability and different sensor properties typically make it almost impossible to employ previously trained classifiers for new data without a significant drop in classification accuracy. Domain adaptation (DA) techniques been proven useful to alleviate that problem. In particular, appearance adaptation techniques may be used to adapt images from a specific dataset in such a way that the generated images have a style that is similar to the images from another dataset. Such techniques are, however, prone to creating artifacts that hinder proper classification of the adapted images. In this work we propose an unsupervised DA approach for change detection tasks, which is based on a particular appearance adaptation method: the Cycle-Consistent Generative Adversarial Network (CycleGAN). Specifically, we extend that method by introducing additional constraints in the training phase of the model components, which make it preserve the semantic structure and class transitions in the adapted images. We evaluate the proposed approach on a deforestation detection application, considering different sites in the Amazon rain-forest and in the Brazilian Cerrado (savanna) using Landsat-8 images. In the experiments, each site corresponds to a domain, and the accuracy of a classifier trained with images and references from one (source) domain is measured in the classification of another (target) domain. The results show that the proposed approach is successful in producing artifact-free adapted images, which can be satisfactory classified by the pre-trained source classifiers. On average, the accuracies achieved in the classification of the adapted images outperformed the baselines (when no adaptation was made) by 7.1% in terms of mean average precision, and 9.1% in terms of F1-Score. To the best of our knowledge, the proposed method is the first unsupervised domain adaptation approach devised for change detection.

KW - Change detection

KW - CycleGAN

KW - Deep learning

KW - Deforestation detection

KW - Domain adaptation

KW - Remote sensing

UR - http://www.scopus.com/inward/record.url?scp=85115024669&partnerID=8YFLogxK

U2 - 10.1016/j.isprsjprs.2021.08.026

DO - 10.1016/j.isprsjprs.2021.08.026

M3 - Article

AN - SCOPUS:85115024669

VL - 181

SP - 113

EP - 128

JO - ISPRS Journal of Photogrammetry and Remote Sensing

JF - ISPRS Journal of Photogrammetry and Remote Sensing

SN - 0924-2716

ER -