Depth cameras enable long term re-identification exploiting 3D information that captures biometric cues such as face and body characteristic lengths. People re-identification is otherwise performed using appearance, thus invalidating any application in which a person may change dress between acquisitions. This is a relevant scenario for home patient monitoring for example. Unfortunately, face and skeleton quality is not always enough to grant a correct recognition. Both features are affected by the subject pose and distance from camera. We propose a model to incorporate a robust skeleton representation with a highly discriminative face feature, weighting samples by their quality. Our method improves rank-1 accuracy especially on short realistic sequences.
Long Term Person Re-Identification from Depth Cameras using Facial and Skeleton Data / Bondi, Enrico; Pala, Pietro; Seidenari, Lorenzo; Berretti, Stefano; Del Bimbo, Alberto. - STAMPA. - 10188 LNCS:(2018), pp. 29-41. (Intervento presentato al convegno 2nd International Workshop on Understanding Human Activities through 3D Sensors (UHA3DS'16) tenutosi a Cancun, Mexico nel 4 December, 2016) [10.1007/978-3-319-91863-1_3].
Long Term Person Re-Identification from Depth Cameras using Facial and Skeleton Data
BONDI, ENRICO;PALA, PIETRO;SEIDENARI, LORENZO;BERRETTI, STEFANO;DEL BIMBO, ALBERTO
2018
Abstract
Depth cameras enable long term re-identification exploiting 3D information that captures biometric cues such as face and body characteristic lengths. People re-identification is otherwise performed using appearance, thus invalidating any application in which a person may change dress between acquisitions. This is a relevant scenario for home patient monitoring for example. Unfortunately, face and skeleton quality is not always enough to grant a correct recognition. Both features are affected by the subject pose and distance from camera. We propose a model to incorporate a robust skeleton representation with a highly discriminative face feature, weighting samples by their quality. Our method improves rank-1 accuracy especially on short realistic sequences.File | Dimensione | Formato | |
---|---|---|---|
icpr_wks16.pdf
Accesso chiuso
Descrizione: articolo principale
Tipologia:
Versione finale referata (Postprint, Accepted manuscript)
Licenza:
Tutti i diritti riservati
Dimensione
2.65 MB
Formato
Adobe PDF
|
2.65 MB | Adobe PDF | Richiedi una copia |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.