A graph-kernel method for re-identification

Luc Brun (1) , Donatello Conte (2) , Pasquale Foggia (2) , Mario Vento (2, 3)

1 Equipe Image - Laboratoire GREYC - UMR6072
2 MIVIA - Intelligent Machines for Video, Image and Audio recognition
3 MIVIA - Machine Intelligence Lab for Video, Image and Audio Processing

Re-identification, that is recognizing that an object appear- ing in a scene is a reoccurrence of an object seen previously by the sys- tem (by the same camera or possibly by a different one) is a challenging problem in video surveillance. In this paper, the problem is addressed using a structural, graph-based representation of the objects of inter- est. A recently proposed graph kernel is adopted for extending to this representation the Principal Component Analyisis (PCA) technique. An experimental evaluation of the method has been performed on two video sequences from the publicly available PETS2009 database.

Mots clés: re identification; Video; graph kernel; re identification.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Luc Brun, Donatello Conte, Pasquale Foggia, Mario Vento. A graph-kernel method for re-identification. International Conference on Image Analysis and Recognition, Jun 2011, Burnaby, Canada. pp.173-182. ⟨hal-00770260⟩