Multimodal Fusion Transformer for Remote Sensing Image Classification

计算机科学 变压器 人工智能 卷积神经网络 CLs上限 编码器 上下文图像分类 模式识别(心理学) 安全性令牌 源代码 机器学习 图像(数学) 量子力学 医学 操作系统 验光服务 物理 电压 计算机安全
作者
Swalpa Kumar Roy,Ankur Deria,Danfeng Hong,Behnood Rasti,Antonio Plaza,Jocelyn Chanussot
出处
期刊:IEEE Transactions on Geoscience and Remote Sensing [Institute of Electrical and Electronics Engineers]
卷期号:61: 1-20 被引量:189
标识
DOI:10.1109/tgrs.2023.3286826
摘要

Vision transformers (ViTs) have been trending in image classification tasks due to their promising performance when compared to convolutional neural networks (CNNs). As a result, many researchers have tried to incorporate ViTs in hyperspectral image (HSI) classification tasks. To achieve satisfactory performance, close to that of CNNs, transformers need fewer parameters. ViTs and other similar transformers use an external classification (CLS) token which is randomly initialized and often fails to generalize well, whereas other sources of multimodal datasets, such as light detection and ranging (LiDAR) offer the potential to improve these models by means of a CLS. In this paper, we introduce a new multimodal fusion transformer (MFT) network which comprises a multihead cross patch attention (mCrossPA) for HSI land-cover classification. Our mCrossPA utilizes other sources of complementary information in addition to the HSI in the transformer encoder to achieve better generalization. The concept of tokenization is used to generate CLS and HSI patch tokens, helping to learn a {distinctive representation} in a reduced and hierarchical feature space. Extensive experiments are carried out on {widely used benchmark} datasets {i.e.,} the University of Houston, Trento, University of Southern Mississippi Gulfpark (MUUFL), and Augsburg. We compare the results of the proposed MFT model with other state-of-the-art transformers, classical CNNs, and conventional classifiers models. The superior performance achieved by the proposed model is due to the use of multihead cross patch attention. The source code will be made available publicly at \url{https://github.com/AnkurDeria/MFT}.}
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
Hongsong发布了新的文献求助10
1秒前
勤恳马里奥完成签到,获得积分0
2秒前
2秒前
yzy发布了新的文献求助10
2秒前
3秒前
3秒前
科目三应助AA采纳,获得10
3秒前
3秒前
Elaine发布了新的文献求助10
3秒前
Elaine发布了新的文献求助10
3秒前
Elaine发布了新的文献求助10
3秒前
Elaine发布了新的文献求助10
3秒前
roy完成签到 ,获得积分10
4秒前
Akashi发布了新的文献求助10
4秒前
李爱国应助茂密的头发采纳,获得10
4秒前
张时婕完成签到 ,获得积分10
4秒前
小胖鱼发布了新的文献求助10
4秒前
不许焦绿o完成签到,获得积分10
5秒前
pcr163应助zhanzhanzhan采纳,获得50
5秒前
5秒前
SweepingMonk应助EthanChan采纳,获得10
5秒前
爆米花应助二豆子0采纳,获得10
6秒前
科研通AI5应助Mian采纳,获得10
6秒前
CodeCraft应助酒九采纳,获得10
6秒前
星辰大海应助不喝可乐采纳,获得10
6秒前
6秒前
7秒前
WJ发布了新的文献求助10
7秒前
JamesPei应助落寞的紫山采纳,获得10
7秒前
平常的不平完成签到,获得积分10
8秒前
系统提示发布了新的文献求助10
8秒前
盛yyyy完成签到,获得积分10
8秒前
合适山河发布了新的文献求助10
9秒前
周星星完成签到 ,获得积分10
9秒前
NexusExplorer应助潦草采纳,获得10
9秒前
ZHEN发布了新的文献求助10
10秒前
艺玲发布了新的文献求助10
11秒前
dddddddio完成签到 ,获得积分10
11秒前
11秒前
gaos发布了新的文献求助10
11秒前
高分求助中
Continuum Thermodynamics and Material Modelling 3000
Production Logging: Theoretical and Interpretive Elements 2700
Social media impact on athlete mental health: #RealityCheck 1020
Ensartinib (Ensacove) for Non-Small Cell Lung Cancer 1000
Unseen Mendieta: The Unpublished Works of Ana Mendieta 1000
Bacterial collagenases and their clinical applications 800
El viaje de una vida: Memorias de María Lecea 800
热门求助领域 (近24小时)
化学 材料科学 生物 医学 工程类 有机化学 生物化学 物理 纳米技术 计算机科学 内科学 化学工程 复合材料 基因 遗传学 物理化学 催化作用 量子力学 光电子学 冶金
热门帖子
关注 科研通微信公众号,转发送积分 3527521
求助须知:如何正确求助?哪些是违规求助? 3107606
关于积分的说明 9286171
捐赠科研通 2805329
什么是DOI,文献DOI怎么找? 1539901
邀请新用户注册赠送积分活动 716827
科研通“疑难数据库(出版商)”最低求助积分说明 709740