Few-shot incremental learning with continual prototype calibration for remote sensing image fine-grained classification

过度拟合 校准 班级(哲学) 计算机科学 相似性(几何) 人工智能 任务(项目管理) 图像(数学) 数据挖掘 一般化 模式识别(心理学) 机器学习 人工神经网络 数学 统计 经济 管理 数学分析
作者
Zining Zhu,Peijin Wang,Wenhui Diao,Jinze Yang,Hongqi Wang,Xian Sun
出处
期刊:Isprs Journal of Photogrammetry and Remote Sensing [Elsevier BV]
卷期号:196: 210-227 被引量:36
标识
DOI:10.1016/j.isprsjprs.2022.12.024
摘要

With the rapid acquisition of remote sensing (RS) data, new categories of objects continue to emerge, and some categories can only obtain a few training samples. Thus, few-shot class-incremental learning (FSCIL) has received intense attention in recent years. Nevertheless, the existing FSCIL methods are difficult to adapt to RS characteristics with high inter-class similarity, large intra-class differences, and complex backgrounds. In this paper, we propose the FSCIL method called continual prototype calibration (CPC), which applies the decoupled learning strategy to cope with the demand to retrain the model when new few-shot categories appear. It can also address the tedious problems in fine-grained RS images by calibrating inter-class and intra-class prototypes. Concretely, considering the severe issue of inter-class confusion, we introduce a prototype separability module (PSM) to update the distribution of inter-class prototypes. In this way, we can distinguish different categories more clearly by weakening their similarities, and the phenomenon of catastrophic forgetting can also be effectively mitigated. Furthermore, to generate more representative and accurate intra-class prototypes, we design a Meta-Network based on foreground enhancement mechanism (FEM). The Meta-Network can boost the learning ability of the model for few-shot data and suppress the occurrence of overfitting by training on task-based data. And the FEM enhances the object features by filtering out redundant background information, facilitating fine-grained classification. Our CPC method is the first FSCIL method applied to optical RS images, and we verify its effectiveness on three datasets. Extensive experiments illustrate that our method achieves state-of-the-art performance in FSCIL of fine-grained RS images. The code is available at https://github.com/ningerhhh/CPC.
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
刚刚
行则将至完成签到,获得积分10
1秒前
初景的应助被柚子采纳,获得20
1秒前
缓慢寒梅发布了新的文献求助20
2秒前
科研通AI6.2的应助被ygj采纳,获得10
2秒前
源代码发布了新的文献求助10
2秒前
jiangyang发布了新的文献求助20
2秒前
3秒前
陶醉发布了新的文献求助10
3秒前
遇见完成签到,获得积分10
3秒前
hrs发布了新的文献求助10
3秒前
3秒前
4秒前
悲凉的素完成签到,获得积分10
4秒前
脑洞疼的应助被谨慎的秋烟采纳,获得100
4秒前
5秒前
6秒前
王珑发布了新的文献求助10
6秒前
yi发布了新的文献求助10
7秒前
陶醉的嘉熙完成签到,获得积分10
7秒前
贾文浩发布了新的文献求助10
10秒前
kk完成签到,获得积分10
10秒前
竹马子发布了新的文献求助10
10秒前
乐乐的应助被帅气诗槐采纳,获得10
10秒前
rhx发布了新的文献求助10
11秒前
粗心的凤完成签到 ,获得积分10
12秒前
12秒前
汉堡包的应助被蓝斯登采纳,获得10
12秒前
hss完成签到,获得积分10
13秒前
晚茶听风完成签到 ,获得积分20
13秒前
橘络发布了新的文献求助10
13秒前
13秒前
费可完成签到,获得积分20
13秒前
pb完成签到,获得积分10
14秒前
14秒前
14秒前
只因完成签到,获得积分10
14秒前
lijiqi完成签到,获得积分10
15秒前
学术垃圾批发商完成签到 ,获得积分10
15秒前
15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
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
plantTFDB 508
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8168144
求助须知:如何正确求助?哪些是违规求助? 9561199
关于积分的说明 21759542
捐赠科研通 7759434
什么是DOI,文献DOI怎么找? 3429948
关于科研通互助平台的介绍 2553742
邀请新用户注册赠送积分活动 2474463