- Skewed t-distribution for Hyperspectral Anomaly Detection based on Autoencoder
Copyright 2021, Koray Kayabol, Ensar Burak Aytekin, Sertac Arisoy, Ercan Engin Kuruoglu
All Rights Reserved
Permission to use, copy, modify, and distribute this software and its documentation for any non-commercial purpose is hereby granted without fee, provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation, and that the name of the author not be used in advertising or publicity pertaining to distribution of the software without specific, written prior permission.
- General information
Thanks for your interest in our work. This is a MATLAB implementation for the Skewed t-distribution for hyperspectral anomaly detection based on autoencoder algorithm. The theoretical detailed of the algorithm can be found in the following paper
Koray Kayabol, Ensar Burak Aytekin, Sertac Arisoy, Ercan Engin Kuruoglu "Skewed t-distribution for Hyperspectral Anomaly Detection based on Autoencoder" IEEE Geoscience and Remote Sensing Letters (GRSL).
If you use this software, you should cite the aforementioned paper in any resulting publication.
If you have any questions about the code, please contact us.
Koray Kayabol via koray.kayabol@gtu.edu.tr,
Ensar Burak Aytekin via burak.aytekin@tubitak.gov.tr,
Sertac Arisoy via sarisoy@gtu.edu.tr .
- Usage
demo matlab code demonstrates of the usage of the code.
mainScript: compute the anomaly detection score for Multivariate Skewed t distribution ,Multivariate Cauchy distribution, Multivariate Jeffrey’s distribution , Multivariate Student’s t distribution and Multivariate Laplace distribution, Multivariate Normal distribution background modelling .
Input:
data : is Real HSI.
map : is anomaly map.
diffim : is Real HSI - Synthesized HSI.
FARValue : is False Alarm Rate for Detection Map
FiltOn : is Filter Selection.
PCAOn : is Principal component analysis ON/OFF Selection.
Output:
Result.AUCScores : is The Area Under the Curve score.
Result.FARScores : is False Alarm Rate score.
Result.elapsedTimeMean : is processing time mean.
Result.dtmap : is detection map.