Detecting small group activities from multimodal observations - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Applied Intelligence Year : 2009

Detecting small group activities from multimodal observations


This article addresses the problem of detecting configurations and activities of small groups of people in an augmented environment. The proposed approach takes a continuous stream of observations coming from differ- ent sensors in the environment as input. The goal is to separate distinct distributions of these observations corre- sponding to distinct group configurations and activities. This article describes an unsupervised method based on the cal- culation of the Jeffrey divergence between histograms over observations. These histograms are generated from adjacent windows of variable size slid from the beginning to the end of a meeting recording. The peaks of the resulting Jeffrey di- vergence curves are detected using successive robust mean estimation. After a merging and filtering process, the re- tained peaks are used to select the best model, i.e. the best allocation of observation distributions for a meeting record- ing. These distinct distributions can be interpreted as distinct segments of group configuration and activity. To evaluate this approach, 5 small group meetings, one seminar and one cocktail party meeting have been recorded. The observations A short version of this article [6] obtained the Best Paper Award of the 3rd IFIP Conference on Artificial Intelligence Applications and Innovations (AIAI) 2006. O. Brdiczka (ﰌ) * J. Maisonnasse * P. Reignier * J.L. Crowley INRIA Rhône-Alpes, 655 avenue de l'Europe, 38334 Saint Ismier Cedex, France e-mail: J. Maisonnasse e-mail: P. Reignier e-mail: J.L. Crowley e-mail: of the small groups meetings and the seminar were gener- ated by a speech activity detector, while the observations of the cocktail party meeting were generated by both the speech activity detector and a visual tracking system. The authors measured the correspondence between detected seg- ments and labeled group configurations and activities. The obtained results are promising, in particular as the method is completely unsupervised.
No file

Dates and versions

hal-00665329 , version 1 (01-02-2012)


  • HAL Id : hal-00665329 , version 1


Oliver Brdiczka, Jérôme Maisonnasse, Patrick Reignier, James L. Crowley. Detecting small group activities from multimodal observations. Applied Intelligence, 2009, 30 (1), pp.47-56. ⟨hal-00665329⟩
123 View
0 Download


Gmail Facebook X LinkedIn More