IntheArt

2026

Deep Learning application for classification of images provided by the SVOM real time GRB trigger

 

Louise Delfosse, Irfu/Dap; Stéphane Schanne, Irfu/Dap

Vendredi 21 aout à 11h30, CEA-Saclay, Orme-des merisiers, salle Cassini

 

The objective of this project is to design a decision-support system for classifying the alerts sent by the ECLAIRs onboard trigger to the ground in real-time over the SVOM VHF network. By automating the classification between astrophysical sources and false alerts using deep learning and computer vision algorithms, the aim of the project is to optimize and reduce the workload of the SVOM Burst Advocates on shift and increase the efficiency of the follow-up observations. A hybrid image classification architecture has been designed, combining, on one hand, an Autoencoder for the extraction of complex features coupled with a Random Forest classifier for filtering, and, on the other hand, a deep neural network ResNet for in-depth analysis. The complementary of both approaches, supervised and unsupervised machine learning, improves the classification reliability, robustness and efficiency.

 

Unsupervised Machine Learning for Fault detection in turn by turn Beam Position Monitors

Jakub Potaczala, Irfu/DACM et Université Paris-Saclay

Jeudi 16 juillet à 13h30, CEA-Saclay, Bat 130, pièce 102

Reporté

Turn-by-turn beam position monitors (TbT-BPMs) are essential for reconstructing the magnetic optics of high-luminosity colliders, yet faulty sensors and measurement noise routinely compromise data quality. As accelerators scale toward facilities like the Future Circular Collider (FCC), with thousands of BPMs distributed across a 91-km ring, manual inspection and hardware redundancy no longer suffice, motivating fully automated, unsupervised diagnostic pipelines. This seminar presents the improvements developed during an internship to a tool under development at CEA in collaboration with LISN at CentraleSupelec. The tool is used to study and compare complementary unsupervised machine learning approaches to TbT-BPM data from SuperKEKB's High and Low Energy Rings (HER/LER). A new pipeline including Isolation Forest, DBSCAN-based clustering coupled with two feature-engineering philosophies (a fully data-driven approach, Time2Feat, and a physics-informed fixed-feature approach) and a 1D convolutional autoencoder. Overall, these methods converge on a shared conclusion: unsupervised, physically interpretable pipelines can match or exceed manual fault curation, offering a scalable foundation for autonomous BPM diagnostics in next-generation accelerators.

 

Les IA Classification of VHF images through Machine Learning

Baptiste Hubert, Irfu/Dap; Stéphane Schanne, Irfu/Dap

Lundi 9 février à 12h, CEA-Saclay, Orme-des merisiers, salle kepler

The SVOM satellite mission, launched in June 2024 is dedicated to Gamma-Ray Burst (GRBs) studies. The ECLAIRs trigger onboard SVOM, which reorients the satellite for GRB follow-up observations, also provides a real-time Alert Sequence for each detected GRB, transmitted to ground over the SVOM VHF receiver network. One of the two trigger algorithms, the Image Trigger (IMT) transmits at the end of each Alert Sequence a Sub-Image centered on the detected source, which is currently used by the trigger team on ground to manually identify a true point-like source from a false alert, by observing the shape in the Sub-Image. We study multiple machine learning algorithms, such as pre-trained Convolutional Neural Networks (CNNs), to perform the automatic classification of the transmitted Sub-Image and rapid identification of False Alerts. This algorithm is intended to be first used on ground at the SVOM French Science Center, and in a later stage may be implemented on board in a future flight software update.

slides, video