Detecting changes in network-structured time series data is of utmost importance in critical applications as diverse as detecting denial of service attacks against online service providers, or monitoring energy and water supplies. We recently devised an online change-point detection algorithm that fully benefits from the recent advances in graph signal processing to exploit the characteristics of the data that lie on irregular supports. The algorithm is able to detect anomalous clusters of activity over time in streaming scalar data measured at each node of the graph in a distributed manner across nodes. A natural extension is to tackle the case where the measurement at each node is vectored value. This paves the way to many applications such as detection of changes in hyperspectral images where a node is associated to each pixel, the graph models spatial dependences between pixels and each node is associated to the corresponding spectrum. The internship will focus on theoretical aspects related to the development of a multivariate detection algorithm, including multivariate graph signal processing, and practical application to change detection in multi temporal hyperspectral images.
To apply for this intership, please contact me with a CV.
Location: Lagrange Lab., Campus Valrose, University Côte d'Azur, Nice (downtown), France