The_final_manuscript_TGRS.pdf (8.54 MB)

Enhanced Feature Pyramid Network with Deep Semantic Embedding for Remote Sensing Scene Classification

Download (8.54 MB)
journal contribution
posted on 07.01.2021, 16:36 by X Wang, S Wang, C Ning, Huiyu Zhou
Recent progress on remote sensing scene classification is substantial, benefiting mostly from the explosive development of convolutional neural networks (CNNs). However, different from the natural images in which the objects occupy most of the space, objects in remote sensing images are usually small and separated. Therefore, there is still a large room for improvement of the vanilla CNNs that extract global image-level features for remote sensing scene classification, ignoring local object-level features. In this paper, we propose a novel remote sensing scene classification method via enhanced feature pyramid network with deep semantic embedding. Our proposed framework extracts multi-scale multi-level features using an enhanced feature pyramid network (EFPN). Then, to leverage the complementary advantages of the multi-level and multi-scale features, we design a deep semantic embedding (DSE) module to generate discriminative features. Third, a feature fusion module, called two-branch deep feature fusion (TDFF), is introduced to aggregate the features at different levels in an effective way. Our method produces state-of-the-art results on two widely used remote sensing scene classification benchmarks, with better effectiveness and accuracy than the existing algorithms. Beyond that, we conduct an exhaustive analysis on the role of each module in the proposed architecture, and the experimental results further verify the merits of the proposed method.

Funding

Wang is supported in part by the Fundamental Research Funds for the Central Universities under Grant B210202077, Six Talents Peak Project of Jiangsu Province under Grant XYDXX-007, Jiangsu Province Government Scholarship for Studying Abroad. H. Zhou is supported in part by Royal Society-Newton Advanced Fellowship under Grant NA160342, and European Union’s Horizon 2020 Research and Innovation Program through the Marie-Sklodowska-Curie under Grant 720325.

History

Author affiliation

School of Informatics

Version

AM (Accepted Manuscript)

Published in

IEEE Transactions on Geoscience and Remote Sensing

Publisher

Institute of Electrical and Electronics Engineers

issn

0196-2892

Acceptance date

11/12/2020

Copyright date

2021

Available date

07/01/2021

Language

en