In this work, we propose an efficient and effective method to recognize human actions based on the estimated 3D positions of skeletal joints in temporal sequences of depth maps. First, the body skeleton is decomposed in a set of kinematic chains, and the position of each joint is expressed in a locally defined reference system, which makes the coordinates invariant to body translations and rotations. A multi-part bag-of-poses approach is then defined, which permits the separate alignment of body parts through a nearest-neighbor classification. Experiments conducted on the MSR Daily Activity dataset show promising results.
Weakly Aligned Multi-Part Bag-of-Poses for Action Recognition from Depth Cameras / L. Seidenari; V. Varano; S. Berretti; A. Del Bimbo; P. Pala. - STAMPA. - 8158:(2013), pp. 446-455. (Intervento presentato al convegno International Workshop on Social Behaviour Analyis (SBA'13) tenutosi a Napoli, Italia nel 10 Settembre 2013) [10.1007/978-3-642-41190-8_48].
Weakly Aligned Multi-Part Bag-of-Poses for Action Recognition from Depth Cameras
SEIDENARI, LORENZO;VARANO, VINCENZO;BERRETTI, STEFANO;DEL BIMBO, ALBERTO;PALA, PIETRO
2013
Abstract
In this work, we propose an efficient and effective method to recognize human actions based on the estimated 3D positions of skeletal joints in temporal sequences of depth maps. First, the body skeleton is decomposed in a set of kinematic chains, and the position of each joint is expressed in a locally defined reference system, which makes the coordinates invariant to body translations and rotations. A multi-part bag-of-poses approach is then defined, which permits the separate alignment of body parts through a nearest-neighbor classification. Experiments conducted on the MSR Daily Activity dataset show promising results.File | Dimensione | Formato | |
---|---|---|---|
iciap13_actionrecognition_10pp.pdf
Accesso chiuso
Tipologia:
Altro
Licenza:
Tutti i diritti riservati
Dimensione
745.89 kB
Formato
Adobe PDF
|
745.89 kB | Adobe PDF | Richiedi una copia |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.