论文信息
论文标题:Anomaly Detection in IR Images of PV Modules using Supervised Contrastive Learning
论文作者:Abhay Rawat, Isha Dua, Saurav Gupta, Rahul Tallamraju
论文来源:Lukas Bommes, Mathis Hoffmann, Claudia Buerhop-Lutz, Tobias Pickel, Jens Hauch, Christoph Brabec, Andreas Maier, Ian Marius Peters
论文地址:download
论文代码:download
视屏讲解:click
1 介绍
从域适应的角度做异常检测;
小记一下,这个方向比较贴合我现在做的东西;
2 方法
2.1 整体框架
2.2 方法简介
在类不平衡源域数据集上使用监督对比训练好一个特征提取器,测试的时候使用同一个特征提取器得到目标域的特征然后使用 KNN 分类器对目标域打标签;(二分类)
3 实验
略;
4 总结
略;
- Contrastive Supervised Detection Learning Anomalycontrastive supervised detection learning mean-shifted contrastive detection anomaly detection learning anomaly survey detection机器anomaly distribution detection anomaly论文 isolation detection anomaly forest discrimination distribution detection anomaly multi-class detection unified anomaly clda semi-supervised contrastive adaptation monocular auxiliary detection learning