Foreground-Aware Relation Network for Geospatial Object Segmentation in High Spatial Resolution Remote Sensing Imagery

By Zhuo Zheng, Yanfei Zhong, Junjue Wang, Ailong Ma. Computer Vision and Pattern Recognition, 2020

325 citations recorded by the source corpus.

Geospatial object segmentation, as a particular semantic segmentation task, always faces with larger-scale variation, larger intra-class variance of background, and foreground-background imbalance in the high spatial resolution (HSR) remote sensing imagery. However, general semantic segmentation methods mainly focus on scale variation in the natural scene, with inadequate consideration of the other two problems that usually happen in the large area earth observation scene. In this paper, we argue that the problems lie on the lack of foreground modeling and propose a foreground-aware relation network (FarSeg) from the perspectives of relation-based and optimization-based foreground modeling, to alleviate the above two problems. From perspective of relation, FarSeg enhances the discrimination of foreground features via foreground-correlated contexts associated by learning foreground-scene relation. Meanwhile, from perspective of optimization, a foreground-aware optimization is proposed to focus on foreground examples and hard examples of background during training for a balanced optimization. The experimental results obtained using a large scale dataset suggest that the proposed method is superior to the state-of-the-art general semantic segmentation methods and achieves a better trade-off between speed and accuracy.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/b569ea0c49ee43437949702edc46eab341ab2eeb

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Dacheng Tao
2. Jing Zhang
3. Bo Du
4. Qiming Zhang
5. Yufei Xu
6. Di Wang
7. Yanfei Zhong
8. Qian Du

In Media Technology

1. Bo Du
2. Qian Du
3. Yanfei Zhong
4. Dacheng Tao
5. Zhuo Zheng
6. Qiming Zhang
7. Jing Zhang
8. Wesley Nunes Gonçalves

Citations in the corpus

• DynamicEarthNet: Daily Multi-Spectral Satellite Dataset for Semantic Change Segmentation (2022, background): “Recent approaches show that encoder-decoder architectures [18, 23] can help to address the foreground-background imbalance of satellite data [22,48].”
• Advancing Plain Vision Transformer Toward Remote Sensing Foundation Model (2022, methodology): “Previous methods mostly utilize ImageNet [27] dataset for pretraining and obtain performance improvement in classification, detection, and segmentation with task-specific designs [28]–[30].”
• LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation (2021, background): “As is shown in Figure 2(c), the background class contains the most pixels with complex samples [29, 53].”
• SuperYOLO: Super Resolution Assisted Object Detection in Multimodal Remote Sensing Imagery (2022, background): “Second, the size of objects in RSI is much smaller, accounting for merely tens of pixels in relation to the complicated and broad backgrounds [9], [10].”
• The Segment Anything Model (SAM) for Remote Sensing Applications: From Zero to One Shot (2023, methodology): “Foreground Segmentation, also known as background subtraction, is a technique primarily used to separate the main subjects or objects of interest (the foreground) from the backdrop (the background) in an image [72, 36].”