BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:6094@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20220404T143000
DTEND;TZID=Europe/Paris:20220404T153000
DTSTAMP:20241120T200911Z
URL:https://www.i2m.univ-amu.fr/evenements/time-series-analysis-of-massive
 -satellite-images-application-to-earth-observation/
SUMMARY:Alexandre CONSTANTIN (GAIA team in GIPSA-lab\, Grenoble): Time-Seri
 es Analysis of Massive Satellite Images: Application to Earth Observation
DESCRIPTION:Alexandre CONSTANTIN: This work takes place in the context of t
 he processing of the Satellite Image Time-Series (SITS) data from Sentinel
 -2 mission\, launched in 2015 by the European Space Agency. It focuses on 
 the classification task\, i.e. land use or land cover maps that can be pro
 duced using spectro-temporal aspect of the Sentinel-2 SITS.\nThe main diff
 iculty is the acquisition noise (clouds\, shadows) combined with the orbit
 al path of the satellites resulting in irregular and unevenly sampled time
 -series. Conventional approaches re-sample time-series to a set of time st
 amps\, then they use machine learning techniques to classify vectors at a 
 large-scale (national scale). The main disadvantage of this two-step proce
 ssing approach is that it increases the number of operations applied to th
 e SITS\, implying a more difficult transition to massive amount of data. T
 o a lower extent\, the re-sampling step may slightly alter the temporal ch
 aracteristics of the data.\nTo tackle this concern\, we introduce a novel 
 model-based approach with the ability to classify irregularly sampled time
 -series based on a mixture of multivariate Gaussian processes using spectr
 al and temporal information from SITS thanks to a Kronecker structure of t
 he covariance operator of the Gaussian process. The model allows jointly t
 he imputation of missing values and is scalable to large-scale data-sets. 
 It is evaluated numerically on simulated data-sets to illustrate the impor
 tance of taking into account spectral correlation and on Sentinel-2 SITS i
 n terms of classification and imputation accuracy and is compared with con
 ventional approaches. Analyses of the results illustrate the relevance of 
 the model and the benefit of using interpretable parametric models.
ATTACH;FMTTYPE=image/jpeg:https://www.i2m.univ-amu.fr/wp-content/uploads/2
 022/04/Alexandre_Constantin.jpg
CATEGORIES:Séminaire,Signal et Apprentissage
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20220327T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR