Spacecraft Recognition Leveraging Knowledge of Space Environment: Simulator, Dataset, Competition Design and Analysis
Mohamed Adel Musallam, Vincent Gaudillière, Enjie Ghorbel, Kassem Al Ismaeil, Marcos Damian Perez, Michel Poucet, Djamila Aouada
Abstract
SPARK represents the first edition of the SPAcecraft Recognition leveraging Knowledge of space environment competition organized by the Interdisciplinary Centre for Security, Reliability and Trust (SnT) in conjunction with the 2021 IEEE International Conference in Image Processing (ICIP 2021). By providing a unique synthetic dataset composed of 150k annotated multi-modal images, SPARK aims at encouraging researchers to develop innovative solutions for space target recognition and detection. This paper introduces the proposed dataset and provides a global analysis of the results obtained for the 17 submissions.
Topics & Concepts
SpacecraftSPARK (programming language)Computer scienceReliability (semiconductor)Space (punctuation)ModalCompetition (biology)Aerospace engineeringEngineeringChemistryPolymer chemistryPhysicsPower (physics)EcologyBiologyProgramming languageOperating systemQuantum mechanicsSpace Satellite Systems and ControlPlanetary Science and Exploration3D Surveying and Cultural Heritage