Activity Recognition with Volume Motion Templates and Histograms of 3D Gradients - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2015

Activity Recognition with Volume Motion Templates and Histograms of 3D Gradients

Résumé

We propose a new method for activity recognition based on a view independent representation of human motion. Robust 3D volume motion templates (VMTs) are calculated from tracklets. View independence is achieved through a rotation with respect to a canonical orientation. From this volumes, features based on 3D gradients are extracted, projected to a codebook and pooled into a bags-of-words model classified with an SVM classifier. Experiments show that the method outperforms the original HoG3D method.
Fichier non déposé

Dates et versions

hal-01147957 , version 1 (03-05-2015)

Identifiants

  • HAL Id : hal-01147957 , version 1

Citer

Emre Dogan, Gonen Eren, Christian Wolf, Atilla Baskurt. Activity Recognition with Volume Motion Templates and Histograms of 3D Gradients. International Conference on Image Processing, Sep 2015, Quebec City, Canada. ⟨hal-01147957⟩
165 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More