Skip to Main content Skip to Navigation
Conference papers

Deep anomaly detection using self-supervised learning: application to time series of cellular data

Abstract : We present a deep self-supervised method for anomaly detection on time series. We apply this methodology to detect anomalies from cellular times series, in particular cell dry mass, obtained in the context of lens-free microscopy.We propose an innovative, self-supervised, two-step method for anomaly detection on time series. As a first step, a representation of the time series is learned thanks to a 1D-convolutional neural network without any labels. Then, the learnedrepresentation is used to feed a threshold anomaly detector. This new self-supervised learning method is tested on an unlabeleddataset of 9100 time series of dry mass and succeeded in detecting abnormal time series with a precision of 96.6%.
Complete list of metadata

https://hal-cea.archives-ouvertes.fr/cea-03605065
Contributor : Contributeur MAP CEA Connect in order to contact the contributor
Submitted on : Thursday, March 10, 2022 - 6:00:05 PM
Last modification on : Friday, April 15, 2022 - 2:07:36 PM

File

ASPAI21.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : cea-03605065, version 2

Citation

Romain Bailly, Marielle Malfante, Cédric Allier, Lamya Ghenim, Jérôme I. Mars. Deep anomaly detection using self-supervised learning: application to time series of cellular data. ASPAI 2021 - 3rd International Conference on Advances in Signal Processing and Artificial Intelligence, Nov 2021, Porto, Portugal. ⟨cea-03605065v2⟩

Share

Metrics

Record views

202

Files downloads

279