Dropout-based Adversarial Training Networks for Remote Sensing Scene Classification
Scene classification in remote sensing (RS) images is a challenging task due to the lack of well labeled data. Recently, deep transfer learning (DTL) has been proposed to handle this task. However, most DTL methods cannot effectively deal with ambiguous features on class boundaries and multi-modal structures of RS data, so their performance is unsatisfactory. To handle the challenges, this letter presents a novel dropout-based adversarial training network for RS scene classification. Specifically, a dropout-based labelclassifier module is designed to reduce the selection of ambiguous features. Then, a dropout-based domain discriminator module is constructed to capture multi-modal structures of RS images so as to achieve fine-grained alignment between cross-domain distributions. Third, a joint distribution of features and labels is built to further enhance the performance. Experiments on seven public RS data sets show that our model outperforms several state-of-the-arts under different conditions.
History
Citation
IEEE Geoscience and Remote Sensing Letters, 2022, in pressAuthor affiliation
School of Computing and Mathematical Sciences, University of LeicesterVersion
- AM (Accepted Manuscript)