Human activity recognition is an important and active field of research having a wide range of application in numerous fields, including ambient assisted living. Although most of the researches are focused on the single user, the ability to recognize two-person interactions is perhaps more important for its social implications. This paper presents a two-person activity recognition system that uses skeleton data extracted from a depth camera. The human actions are encoded using a set of a few basic postures obtained with an unsupervised clustering approach. Multiclass Support Vector Machines (SVMs) are used to build models on the training set, while the X-means algorithm is employed to dynamically find the optimal number of clusters for each sample during the classification phase. The system is evaluated on the ISR-UoL and SBU datasets, reaching an overall accuracy of 0.87 and 0.88 respectively. Although the results show that the performances of the system are comparable with the state-of-the-art, recognition improvements are obtained with the activities related to health-care environments, showing promise for applications in the assisted-living realm.

Two-person Activity Recognition using Skeleton Data / MANZI, Alessandro; FIORINI, Laura; LIMOSANI, Raffaele; DARIO, Paolo; CAVALLO, Filippo. - In: IET COMPUTER VISION. - ISSN 1751-9632. - STAMPA. - 12:(2018), pp. -27. [10.1049/iet-cvi.2017.0118]

Two-person Activity Recognition using Skeleton Data

FIORINI, Laura;CAVALLO, Filippo
2018

Abstract

Human activity recognition is an important and active field of research having a wide range of application in numerous fields, including ambient assisted living. Although most of the researches are focused on the single user, the ability to recognize two-person interactions is perhaps more important for its social implications. This paper presents a two-person activity recognition system that uses skeleton data extracted from a depth camera. The human actions are encoded using a set of a few basic postures obtained with an unsupervised clustering approach. Multiclass Support Vector Machines (SVMs) are used to build models on the training set, while the X-means algorithm is employed to dynamically find the optimal number of clusters for each sample during the classification phase. The system is evaluated on the ISR-UoL and SBU datasets, reaching an overall accuracy of 0.87 and 0.88 respectively. Although the results show that the performances of the system are comparable with the state-of-the-art, recognition improvements are obtained with the activities related to health-care environments, showing promise for applications in the assisted-living realm.
2018
12
27
MANZI, Alessandro; FIORINI, Laura; LIMOSANI, Raffaele; DARIO, Paolo; CAVALLO, Filippo
File in questo prodotto:
File Dimensione Formato  
IP032 - Two-person Activity Recognition.pdf

Accesso chiuso

Dimensione 3.2 MB
Formato Adobe PDF
3.2 MB Adobe PDF   Richiedi una copia

I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/1210786
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 29
  • ???jsp.display-item.citation.isi??? 27
social impact