Trajectory Based Activity Discovery - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2010

Trajectory Based Activity Discovery


This paper proposes a framework to discover activities in an unsupervised manner, and add semantics with minimal supervision. The framework uses basic trajectory information as input and goes up to video interpretation. The work reduces the gap between low-level information and semantic interpretation, building an intermediate layer composed of Primitive Events. The proposed representation for primitive events aims at capturing small meaningful motions over the scene with the advantage of being learnt in an unsupervised manner. We propose the discovery of an activity using these Primitive Events as the main descriptors. The activity discovery is done using only real tracking data. Semantics are added to the discovered activities and the recognition of activities (e.g., “Cooking”, “Eating”) can be automatically done with new datasets. Finally we validate the descriptors by discovering and recognizing activities in a home care application dataset.
Fichier principal
Vignette du fichier
AVSS10-PBT.pdf (8.19 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

inria-00504634 , version 1 (20-07-2010)


  • HAL Id : inria-00504634 , version 1


Guido Pusiol, François Bremond, Monique Thonnat. Trajectory Based Activity Discovery. 7th IEEE International Conference on Advanced Video and Signal-Based Surveillance, Aug 2010, Boston, United States. ⟨inria-00504634⟩


82 View
105 Download


Gmail Facebook X LinkedIn More