DOCNet: Dual-Domain Optimized Class-Aware Network for Remote Sensing Image Segmentation
- 5 January 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 21 (1545598X) , 1-5
Abstract
The spatial attention mechanism has been frequently employed for the semantic segmentation of remote sensing images, given its renowned capability to model long-range dependencies. As remote sensing images often exhibit intricate backgrounds, significant intraclass variability, and a foreground-background imbalance, spatial attention mechanism-based methods somehow tend to introduce an extensive amount of background context through intensive affinity operations, causing unsatisfactory segmentation outcomes. While several class-aware methods attempt to attenuate the interference of background context by generating class representations as representative features, they still encounter challenges related to independent correlation calculation and single-confidence scale class representations. We introduce a dual-domain optimized class-aware network designed to address these challenges. In the semantic domain, we use category confidence as a scaling criterion to derive class representations at multiple confidence scales, effectively modeling pixel-class relationships. In the spatial domain, we leverage pixel-class relationships and their consensus to enhance relevant correlations while suppressing erroneous ones. Experimental results on three datasets demonstrate that the proposed method surpasses previous state-of-the-art ones for remote sensing image segmentation. Code is available at https://github.com/xwmaxwma/rssegmentation .Keywords
Funding Information
- National Natural Science Foundation of China (62202421)
- Zhejiang Provincial Key Research and Development Program (2021C01031)
- Ningbo Yongjiang Talent Introduction Program (2021A-157-G)
- Public Welfare Science and Technology Plan of Ningbo City (2022S125)
This publication has 27 references indexed in Scilit:
- A ConvNet for the 2020s2022
- MetaFormer is Actually What You Need for Vision2022
- Lake water body extraction of optical remote sensing images based on semantic segmentationApplied Intelligence, 2022
- Foreground-Aware Relation Network for Geospatial Object Segmentation in High Spatial Resolution Remote Sensing Imagery2020
- Deep learning in environmental remote sensing: Achievements and challengesRemote Sensing of Environment, 2020
- Dual Attention Network for Scene Segmentation2019
- CBAM: Convolutional Block Attention Module2018
- Non-local Neural Networks2018
- Pyramid Scene Parsing Network2017
- Mapping urbanization dynamics at regional and global scales using multi-temporal DMSP/OLS nighttime light dataRemote Sensing of Environment, 2011