三巴:远程传感图像的语义细分与状态空间模型
Qinfeng Zhu1,2, Yuanzhi Cai3, Yuan Fang1
1Department of Civil Engineering, Xi'an Jiaotong-Liverpool University, Suzhou, 215123, China.
Heliyon
|October 14, 2024
概括
三巴是一个新的语义细分框架,有效地解决了高分辨率遥感图像的挑战. 它利用状态空间模型 (SSM) 来实现与现有方法相比更高的性能.
科学领域:
- 计算机视觉 计算机视觉
- 遥感 遥感 遥感 遥感
- 深度学习 (Deep Learning) 是一种深度学习.
背景情况:
- 高分辨率的遥感图像对传统的语义细分网络,如卷积神经网络 (CNN) 和视觉转换器 (ViT) 提出了重大挑战.
- 对于高分辨率数据来说,CNN的受体场有限,而ViT则在长序列上扎.
- 国家空间模型 (SSM) 提供了一种有效的方法来捕获全球语义信息.
研究的目的:
- 介绍Samba,一种新的语义细分框架,旨在用于高分辨率遥感图像.
- 为了证明SSM在细分复杂的遥感数据集中的有效性.
- 在这个领域,为基于Mamba的技术建立一个新的性能基准.
主要方法:
- 建议使用编码器-解码器架构的Samba框架.
- 利用多个Samba块,灵感来自Mamba网络,作为多级特征提取的编码器.
- 使用UperNet作为语义细分的解码器.
主要成果:
- 在LoveDA,ISPRS Vaihingen和ISPRS Potsdam数据集上,Samba取得了无与伦比的表现.
- 使用mIoU和mF1指标进行评估,优于顶级基于CNN和基于ViT的方法.
- 展示了SSM在遥感图像中的语义细分的首次成功应用.
结论:
- 三巴显著推进了用于高分辨率遥感数据的语义细分.
- 该框架为远程传感图像分析中的基于Mamba的方法设定了新的标准.
- 该研究强调了SSM在复杂的计算机视觉任务中的潜力.
更多相关视频
08:25Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
Published on: May 7, 2019
8.9K
12:26Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling SAHM
Published on: October 11, 2016
13.3K
相关概念视频
01:27State Space Representation
168
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
Consider an RLC circuit, a...
168
01:21State Space to Transfer Function
176
The conversion of state-space representation to a transfer function is a fundamental process in system analysis. It provides a method for transitioning from a time-domain description to a frequency-domain representation, which is crucial for simplifying the analysis and design of control systems.
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
176
01:23Transfer Function to State Space
201
State-space representation is a powerful tool for simulating physical systems on digital computers, necessitating the conversion of the transfer function into state-space form. Consider an nth-order linear differential equation with constant coefficients, like those encountered in an RLC circuit. The state variables are selected as the output and its n−1 derivatives. Differentiating these variables and substituting them back into the original equation produces the state equations.
In an...
In an...
201