Titre :

Geometric Deep Neural Network using Rigid and Non-Rigid Transformations for Human Action Recognition

Auteur(s) :

Friji, Rasha [Auteur]
Centre de Recherche Réseau Image SysTème Architecture et MuLtimédia [CRISTAL]
Drira, Hassen [Auteur]
Centre for Digital Systems [CERI SN]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Chaieb, Faten [Auteur]
École d'ingénieur généraliste en informatique et technologies du numérique [EFREI]
Kchok, Hamza [Auteur]
Institut National des Sciences Appliquées et de Technologie [Tunis] [INSAT]
Kurtek, Sebastian [Auteur]

Titre de la manifestation scientifique :

International Conference in Computer Vision

Ville :

Visio

Pays :

France

Date de début de la manifestation scientifique :

2021-10-11

Discipline(s) HAL :

Sciences de l'Homme et Société/Sciences de l'information et de la communication
Informatique [cs]/Vision par ordinateur et reconnaissance de formes [cs.CV]

Résumé en anglais : [en]

Deep Learning architectures, albeit successful in most computer vision tasks, were designed for data with an underlying Euclidean structure, which is not usually fulfilled since pre-processed data may lie on a non-linear ...
Lire la suite >Deep Learning architectures, albeit successful in most computer vision tasks, were designed for data with an underlying Euclidean structure, which is not usually fulfilled since pre-processed data may lie on a non-linear space. In this paper, we propose a geometry aware deep learning approach using rigid and non rigid transformation optimization for skeleton-based action recognition. Skeleton sequences are first modeled as trajectories on Kendall's shape space and then mapped to the linear tangent space. The resulting structured data are then fed to a deep learning architecture, which includes a layer that optimizes over rigid and non rigid transformations of the 3D skeletons, followed by a CNN-LSTM network. The assessment on two large scale skeleton datasets, namely NTU-RGB+D and NTU-RGB+D 120, has proven that the proposed approach outperforms existing geometric deep learning methods and exceeds recently published approaches with respect to the majority of configurations.Lire moins >

Langue :

Anglais

Comité de lecture :

Oui

Audience :

Internationale

Vulgarisation :

Non

Collections :

• Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL

Date de dépôt :

2021-12-08T02:01:07Z