FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952852611> ?p ?o ?g. }

• W2952852611 abstract "RESUME : La classification automatique de plusieurs observations en mouvement, utilisant des radars de surveillance au sol, a fait l’objet de beaucoup d’attention en recherche. Le probleme de la classification automatique sans assistance humaine reste un defi pour les systemes radar modernes. En particulier, la distinction entre les petits drones et les oiseaux est un nouveau defi auquel les systemes precedents n’avaient pas a faire face. La classification est souvent effectuee hors ligne, a l’aide de techniques d’apprentissage supervisees, mais la reconnaissance de classes invisibles reste difficile. En regle generale, les methodes de classification en lignenecessitent un niveau eleve d’informations et prennent beaucoup de temps. Dans ce travail, nous decrivons et analysons un modele de classification simple pour une formalisation du probleme, que nous appelons la reconnaissance de cibles de radar avec leurres. De plus, nous evaluons les performances du modele propose a la fois sur des donnees simulees et sur un jeu de donnees radar reel. Les resultats des experiences montrent que le modele propose offre une meilleure performance que les methodes actuelles.----------ABSTRACT : Automatic classification of multiple moving targets, using ground surveillance radars has received a lot of attention in radar technology research. The problem of automatic classification without human assistance remains a challenge for modern radar systems. In particular, distinguishing between small drones and birds is a novel challenging problem that older systemsdid not have to face. The classification is often performed offline, using supervised learning techniques, but recognition of unseen classes remains difficult. Typically, online classification methods requires fine grained information and are very time consuming. This study aims to develop a new method for recognizing the observations from unknown classes. In this work, we describe and analyze a new model for a formalization of the problem, which we call online target recognition with decoys. The proposed model is an extension of Bayesian quadratic discriminant analysis for classification with additional abilities to recognizeobjects from unknown classes, called decoys. We aim to study how effectively the new model can predict the class of the new observations and predict the noise objects as decoys. Furthermore, we evaluate the performance of the proposed model on simulated data and on a real radar dataset, and compare it to quadratic discriminant analysis and support vector machine methods. Experiment results show thatthe proposed model improve over these baselines." @default.
• W2952852611 created "2019-06-27" @default.
• W2952852611 creator A5067174259 @default.
• W2952852611 date "2019-02-01" @default.
• W2952852611 modified "2023-09-26" @default.
• W2952852611 title "Online Radar Target Recognition with Decoys" @default.
• W2952852611 hasPublicationYear "2019" @default.
• W2952852611 type Work @default.
• W2952852611 sameAs 2952852611 @default.
• W2952852611 citedByCount "0" @default.
• W2952852611 crossrefType "journal-article" @default.
• W2952852611 hasAuthorship W2952852611A5067174259 @default.
• W2952852611 hasConcept C138885662 @default.
• W2952852611 hasConcept C154945302 @default.
• W2952852611 hasConcept C15708023 @default.
• W2952852611 hasConcept C17744445 @default.
• W2952852611 hasConcept C2917558 @default.
• W2952852611 hasConcept C41008148 @default.
• W2952852611 hasConcept C554190296 @default.
• W2952852611 hasConcept C76155785 @default.
• W2952852611 hasConceptScore W2952852611C138885662 @default.
• W2952852611 hasConceptScore W2952852611C154945302 @default.
• W2952852611 hasConceptScore W2952852611C15708023 @default.
• W2952852611 hasConceptScore W2952852611C17744445 @default.
• W2952852611 hasConceptScore W2952852611C2917558 @default.
• W2952852611 hasConceptScore W2952852611C41008148 @default.
• W2952852611 hasConceptScore W2952852611C554190296 @default.
• W2952852611 hasConceptScore W2952852611C76155785 @default.
• W2952852611 hasLocation W29528526111 @default.
• W2952852611 hasOpenAccess W2952852611 @default.
• W2952852611 hasPrimaryLocation W29528526111 @default.
• W2952852611 hasRelatedWork W106724427 @default.
• W2952852611 hasRelatedWork W2094252110 @default.
• W2952852611 hasRelatedWork W2166675213 @default.
• W2952852611 hasRelatedWork W2223286428 @default.
• W2952852611 hasRelatedWork W2406008833 @default.
• W2952852611 hasRelatedWork W2417455953 @default.
• W2952852611 hasRelatedWork W2587209668 @default.
• W2952852611 hasRelatedWork W2689651934 @default.
• W2952852611 hasRelatedWork W2785651640 @default.
• W2952852611 hasRelatedWork W2786213716 @default.
• W2952852611 hasRelatedWork W2803699176 @default.
• W2952852611 hasRelatedWork W2943615880 @default.
• W2952852611 hasRelatedWork W2951437415 @default.
• W2952852611 hasRelatedWork W3083200146 @default.
• W2952852611 hasRelatedWork W3106355232 @default.
• W2952852611 hasRelatedWork W3127672287 @default.
• W2952852611 hasRelatedWork W3138942052 @default.
• W2952852611 hasRelatedWork W3153141521 @default.
• W2952852611 hasRelatedWork W3197697977 @default.
• W2952852611 hasRelatedWork W77992279 @default.
• W2952852611 isParatext "false" @default.
• W2952852611 isRetracted "false" @default.
• W2952852611 magId "2952852611" @default.
• W2952852611 workType "article" @default.