Ent?te

Logo du LIFL

Depuis le 1er janvier 2015 le LIFL et le LAGIS forment le laboratoire CRIStAL

  1. Doctoral studies

Thesis of

Rim Slama

Monday 6 October 2014
TELECOM Lille

Geometric Approaches for 3D Human Motion Analysis: Application to Action Recognition and Retrieval

M. Edmond Boyer, Directeur de Recherches, INRIA Grenoble Rhône-Alpes (rapporteur) Mme. Rita Cucchiara, Professeur, University of Modena and Reggio Emilia, Italy (rapporteur) Mme Saida Bouakaz, Professeur, Université Claude Bernard Lyon 1 (examinateur) M. Hubert Cardot, Professeur, Université François Rabelais, Tours (examinateur) M. Olivier Colot, Professeur, Université Lille 1 (examinateur) M. Alain Trouvé, Professeur, École Normale Supérieure de Cachan (examinateur) M. Mohamed Daoudi, Professeur, Institut Mines-Télécom (directeur) M. Hazem Wannous, Maître de conférences, Université Lille 1 (encadrant)

In this thesis, we focus on the development of adequate geometric frameworks in order to model and compare accurately human motion acquired from 3D sensors. In the first framework, we address the problem of pose/motion retrieval in full 3D reconstructed sequences. The human shape representation is formulated using Extremal Human Curve (EHC) descriptor extracted from the body surface. It allows efficient shape to shape comparison taking benefits from Riemannian geometry in the open curve shape space. As each human pose represented by this descriptor is viewed as a point in the shape space, we propose to model the motion sequence by a trajectory on this space. Dynamic Time Warping in the feature vector space is then used to compare different motions. In the second framework, we propose a solution for action and gesture recognition from both skeleton and depth data acquired by low cost cameras such as Microsoft Kinect. The action sequence is represented by a dynamical system whose observability matrix is characterized as an element of a Grassmann manifold. Thus, recognition problem is reformulated as a point classification on this manifold. Here, a new learning algorithm based on the notion of tangent spaces is proposed to improve recognition task. Performances of our approach on several benchmarks show high recognition accuracy with low latency.

Ours

UMR 8022 - Laboratoire d'Informatique Fondamentale de Lille - Copyright © 2012 Sophie TISON - Crédits & Mentions légales

Page respectant XHTML et CSS.

Pour tout commentaire / Comments and remarks : webmaster