Litcius/Paper detail

Skewed <i>t</i> -Distribution for Hyperspectral Anomaly Detection Based on Autoencoder

Koray Kayabol, Ensar Burak Aytekin, Sertaç Arısoy, Erçan E. Kuruoğlu

2021IEEE Geoscience and Remote Sensing Letters15 citationsDOI

Abstract

We propose multivariate skewed <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${t}$ </tex-math></inline-formula> -distribution (MVSkt) for hyperspectral anomaly detection (AD). The proposed distribution model is able to increase the detection performance of autoencoder (AE)-based anomaly detectors. In the proposed method, the reconstruction error of a deep AE is modeled with a skewed <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${t}$ </tex-math></inline-formula> -distribution. The deep AE network is trained based on adversarial learning strategy by feeding its input with the hyperspectral data cubes. The parameters of the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${t}$ </tex-math></inline-formula> -distribution model are estimated using variational Bayesian approach. We define an MVSkt-based detection rule for pixel-wise AD. We compare our proposed method with those based on the multivariate normal (MVN) distribution and the robust MVN variance–mean mixture distributions on real hyperspectral datasets. The experimental results show that the proposed approach outperforms other detectors in the benchmark.

Topics & Concepts

AutoencoderHyperspectral imagingAnomaly detectionArtificial intelligenceNotationMathematicsPattern recognition (psychology)PixelAlgorithmComputer scienceDeep learningArithmeticRemote-Sensing Image ClassificationRemote Sensing and Land UseAdvanced Chemical Sensor Technologies
Skewed <i>t</i> -Distribution for Hyperspectral Anomaly Detection Based on Autoencoder | Litcius