Anomaly Detection in Mixed Time-Series Using A Convolutional Sparse Representation With Application To Spacecraft Health Monitoring

Barbara Pilastre; Gustavo Silva; Loïc Boussouf; Stéphane D'Escrivan; Paul Rodríguez; Jean-Yves Tourneret

ICASSP 2020accepted0 citations

Anomaly Detection in Mixed Time-Series Using A Convolutional Sparse Representation With Application To Spacecraft Health Monitoring

Barbara Pilastre, Gustavo Silva, Loïc Boussouf, Stéphane D'Escrivan, Paul Rodríguez, Jean-Yves Tourneret

Abstract

This paper introduces a convolutional sparse model for anomaly detection in mixed continuous and discrete data. This model, referred to as C-ADDICT, builds upon the experiences of our previous ADDICT algorithm. It can handle discrete and continuous data jointly, is intrinsically shift-invariant, and crucially, it encodes each input signal (either continuous or discrete) from a joint activation and uniform combinations of filters, allowing the correlation across the input signals to be captured. The performance of C-ADDICT, is evaluated on a representative dataset composed of real spacecraft telemetries with an available ground-truth, providing promising results.

BibTeX

@inproceedings{icassp2020_anomalydetection,
  title = {Anomaly Detection in Mixed Time-Series Using A Convolutional Sparse Representation With Application To Spacecraft Health Monitoring},
  author = {Barbara Pilastre and Gustavo Silva and Loïc Boussouf and Stéphane D'Escrivan and Paul Rodríguez and Jean-Yves Tourneret},
  booktitle = {ICASSP 2020},
  year = {2020}
}