Articles
200
Tags
38
Categories
0
Home
Archives
Tags
Categories
List
Music
Movie
Link
About
ALTNT's Hexo Blog
Home
Archives
Tags
Categories
List
Music
Movie
Link
About
Deep learning with multi-scale temporal hybrid structure for robust crop mapping
Created
2025-09-16
|
Updated
2025-10-22
|
Post Views:
Hey, password is required here.
Author:
ALTNT
Link:
http://blog.705553939.xyz/2025/09/16/crop_classification/2024-IF12.2-ISPRS-Multi-scale-Temporal-Transformer-Conv/
Copyright Notice:
All articles in this blog are licensed under
CC BY-NC-SA 4.0
unless stating additionally.
crop classification
Previous
周报2025年9月21日
Next
Learning the Optimal Stopping for Early Classification within Finite Horizons via Sequential Probability Ratio Test
Related Articles
2024-10-07
周报10月 7 日
2024-09-12
A Deep Learning Framework under Attention Mechanism for Wheat Yield Estimation Using Remotely Sensed Indices in the Guanzhong Plain
2024-10-08
Temporal Convolutional Neural Network for the Classification of Satellite Image
2024-10-27
Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Coding
2025-06-19
Climate-adaptive emergency crop monitoring in inaccessible regions with satellite imagery
2024-10-07
记录如何生产 CACM 数据集过程
ALTNT
Articles
200
Tags
38
Categories
0
Follow Me
Announcement
This is my Blog
Contents
1.
Deep learning with multi-scale temporal hybrid structure for robust crop mapping
1.1.
一、研究背景与意义
1.2.
二、核心方法:多尺度时间Transformer - Conv网络(Ms - TTC)
1.2.1.
(一)设计目标
1.2.2.
(二)网络结构
1.3.
三、实验数据:FranceCrops数据集
1.4.
四、实验设置与结果
1.4.1.
(一)对比模型
1.4.2.
(二)实验参数
1.4.3.
(三)评价指标
1.4.4.
(四)实验结果
1.5.
五、模型解释与分析
1.5.1.
(一)模型处理云观测的机制(基于梯度反向传播的特征重要性分析)
1.5.2.
(二)Ms - TTC处理时间信号的机制
1.6.
六、研究结论
1.7.
摘要
1.8.
引言
1.9.
2. 研究方法
1.9.1.
2.1 时间编码器(看起来就是在标准的 transformer 结构基础上改了下前馈层结构啊)
1.9.1.1.
2.1.1 全局感知块
1.9.1.2.
2.1.2 局部关联块
1.9.2.
2.2 多尺度融合模块
1.9.2.1.
2.2.1 基于MsConv的多尺度特征图生成
1.9.2.2.
2.2.2 融合块
1.9.3.
2.3 输出模块
1.10.
3. FranceCrops数据集
1.11.
3.1 作物标签
1.12.
3.2 卫星数据
1.13.
4. 实验
1.13.1.
4.1 与其他模型的对比
1.13.2.
4.2 实验设置
1.13.2.1.
4.2.1 传统浅层模型(RF与SVM)设置
1.13.2.2.
4.2.2 深度学习模型设置
1.14.
5. 实验结果
1.14.1.
5.1 定量评估
1.14.2.
5.2 定性评估
1.14.3.
5.3 消融实验
1.14.4.
5.4 模型鲁棒性与泛化性能
1.14.5.
5.5 抗云鲁棒性
1.15.
6. 讨论
1.15.1.
6.1 不同模型处理含云观测数据的机制解读
1.15.2.
6.2 Ms-TTC模型处理时间信号的机制解读
2.
7. 结论
2.0.1.
并行架构的定义
Recent Post
周报2025年12月7日
2025-12-08
周报2025年11月30日
2025-11-30
LP-FT
2025-11-23
周报2025年11月22日
2025-11-22
In-season crop progress in unsurveyed regions using networks trained on synthetic data
2025-11-16