Fine-Tuning and High-Frequency Enhanced Multi-Scale Segment Anything Model for Remote Sensing Image Semantic Segmentation

Dong Yuqing; Liu Dongsheng; Zhou Xinxin; Fu Junjie; Huang Zhiming; 黄智铭

doi:10.11834/jrs.20266146

Views:101Downloads:17CSCD:0

Export
Share
Collection
Album

Fine-Tuning and High-Frequency Enhanced Multi-Scale Segment Anything Model for Remote Sensing Image Semantic Segmentation
“SAM is widely used in various downstream visual tasks due to its excellent performance and generalization ability. The research team has established the MSHF-SAM model to address the limitations of its migration to remote sensing image semantic segmentation tasks, and explored the fusion mechanism of high-frequency enhancement and efficient parameter fine-tuning, providing a solution to the problem of cross domain adaptation of visual large models.”
- role：First author第一作者
- Affiliation：
  School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
- Email：202412640323@nuist.edu.cn
- Introduction：董雨晴，研究方向为高分辨率遥感语义变化检测。E-mail：202412640323@nuist.edu.cn
Dong Yuqing
1 * ,
- role：Corresponding author通信作者
- Affiliation：
  School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
- Email：ldsh1201@163.com
- Introduction：刘东升，研究方向为遥感图像处理、地理空间智能。E-mail：ldsh1201@163.com
Liu Dongsheng
1 * ,
- Affiliation：
  Institute of National Territory Spatial Intelligent Perception, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
Zhou Xinxin
2 * ,
- Affiliation：
  Beijing Supermap Software Co., Ltd., Beijing 100015, China
Fu Junjie
3 * ,
- Affiliation：
  School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
Huang Zhiming
1 * ,
- Affiliation：
  School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
黄智铭
1 *
Pages: 1-13(2026)
DOI：10.11834/jrs.20266146
Accepted：

Scan QR Code

Quote
Translate The Full Text
DOI：
DOI：

Full Text(HTML)
Figs( 7 ) Tabs( 5 )
References
Publication Info
Metrics

Alert me when the article has been cited

Submit

Fine-Tuning and High-Frequency Enhanced Multi-Scale Segment Anything Model for Remote Sensing Image Semantic Segmentation

Dong Yuqing

Liu Dongsheng

Zhou Xinxin

Fu Junjie

Huang Zhiming

黄智铭

DOI：10.11834/jrs.20266146

Full Text(HTML)

Figs( 7 ) Tabs( 5 )

References

Publication Info

Metrics

摘要

Abstract

关键词

Keywords

references