Remote Sensing Image Change Detection With Transformers

计算机科学 编码器 判别式 变压器 变更检测 人工智能 特征提取 语义特征 模式识别(心理学) 直觉 特征(语言学) 深度学习 卷积码 计算复杂性理论 计算机视觉 目标检测 数据挖掘 自编码 卷积神经网络 数据建模 语义数据模型 遥感 棱锥(几何) 高光谱成像 图像处理 稳健性(进化)
作者
Hao Chen,Zipeng Qi,Zhenwei Shi
出处
期刊:IEEE Transactions on Geoscience and Remote Sensing [Institute of Electrical and Electronics Engineers]
卷期号:60: 1-14 被引量:1007
标识
DOI:10.1109/tgrs.2021.3095166
摘要

Modern change detection (CD) has achieved remarkable success by the powerful discriminative ability of deep convolutions. However, high-resolution remote sensing CD remains challenging due to the complexity of objects in the scene. Objects with the same semantic concept may show distinct spectral characteristics at different times and spatial locations. Most recent CD pipelines using pure convolutions are still struggling to relate long-range concepts in space-time. Nonlocal self-attention approaches show promising performance via modeling dense relationships among pixels, yet are computationally inefficient. Here, we propose a bitemporal image transformer (BIT) to efficiently and effectively model contexts within the spatial-temporal domain. Our intuition is that the high-level concepts of the change of interest can be represented by a few visual words, that is, semantic tokens. To achieve this, we express the bitemporal image into a few tokens and use a transformer encoder to model contexts in the compact token-based space-time. The learned context-rich tokens are then fed back to the pixel-space for refining the original features via a transformer decoder. We incorporate BIT in a deep feature differencing-based CD framework. Extensive experiments on three CD datasets demonstrate the effectiveness and efficiency of the proposed method. Notably, our BIT-based model significantly outperforms the purely convolutional baseline using only three times lower computational costs and model parameters. Based on a naive backbone (ResNet18) without sophisticated structures (e.g., feature pyramid network (FPN) and UNet), our model surpasses several state-of-the-art CD methods, including better than four recent attention-based methods in terms of efficiency and accuracy. Our code is available at https://github.com/justchenhao/BIT_CD.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
RRhhh发布了新的文献求助10
1秒前
ASHUN完成签到,获得积分10
1秒前
小范发布了新的文献求助10
2秒前
爱格儿发布了新的文献求助10
2秒前
xinxin发布了新的文献求助30
2秒前
Lynn完成签到 ,获得积分10
3秒前
3秒前
4秒前
溯7发布了新的文献求助10
5秒前
直率小之完成签到,获得积分20
5秒前
寻我应助Isaiah采纳,获得10
5秒前
xu完成签到,获得积分10
6秒前
李健的小迷弟应助Liar采纳,获得10
6秒前
Lucas应助TJAMO采纳,获得10
7秒前
万能图书馆应助NNUsusan采纳,获得10
8秒前
无极微光应助NNUsusan采纳,获得20
8秒前
王泰一发布了新的文献求助10
9秒前
刘羽萱发布了新的文献求助10
9秒前
10秒前
11秒前
Hello应助。。。采纳,获得10
12秒前
在水一方应助精明金毛采纳,获得10
13秒前
千朝词完成签到,获得积分20
15秒前
15秒前
哒哒哒应助小范采纳,获得10
15秒前
怡然的凌兰应助云云采纳,获得10
16秒前
香蕉觅云应助云云采纳,获得10
16秒前
OsamaKareem应助科研通管家采纳,获得20
16秒前
脑洞疼应助科研通管家采纳,获得10
16秒前
orixero应助科研通管家采纳,获得10
16秒前
充电宝应助科研通管家采纳,获得30
16秒前
思源应助科研通管家采纳,获得10
16秒前
可爱多应助科研通管家采纳,获得20
16秒前
大模型应助科研通管家采纳,获得10
16秒前
16秒前
CodeCraft应助科研通管家采纳,获得10
16秒前
Lucas应助科研通管家采纳,获得10
16秒前
16秒前
SciGPT应助科研通管家采纳,获得10
16秒前
OsamaKareem应助科研通管家采纳,获得20
16秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Les Mantodea de Guyane Insecta, Polyneoptera 2000
Leading Academic-Practice Partnerships in Nursing and Healthcare: A Paradigm for Change 800
基于非线性光纤环形镜的全保偏锁模激光器研究-上海科技大学 800
Pulse width control of a 3-phase inverter with non sinusoidal phase voltages 777
Signals, Systems, and Signal Processing 610
Research Methods for Business: A Skill Building Approach, 9th Edition 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 工程类 有机化学 化学工程 生物化学 计算机科学 物理 内科学 复合材料 催化作用 物理化学 光电子学 电极 细胞生物学 基因 无机化学
热门帖子
关注 科研通微信公众号,转发送积分 6409791
求助须知:如何正确求助?哪些是违规求助? 8228982
关于积分的说明 17459389
捐赠科研通 5462770
什么是DOI,文献DOI怎么找? 2886436
邀请新用户注册赠送积分活动 1862919
关于科研通互助平台的介绍 1702279