Show simple item record

Mid-level features and spatio-temporal context for activity recognition

dc.contributor.authorYuan, Fei
dc.contributor.authorXia, Gui-Song
dc.contributor.authorSahbi, Hichem
HAL ID: 742003
dc.contributor.authorPrinet, Véronique
dc.date.accessioned2013-09-05T11:06:03Z
dc.date.available2013-09-05T11:06:03Z
dc.date.issued2012
dc.identifier.urihttps://basepub.dauphine.fr/handle/123456789/11626
dc.language.isoenen
dc.subjectActivity recognitionen
dc.subjectMid-level featuresen
dc.subjectActivity componentsen
dc.subjectSpatio-temporal context kernelsen
dc.subject.ddc621.3en
dc.titleMid-level features and spatio-temporal context for activity recognitionen
dc.typeArticle accepté pour publication ou publié
dc.description.abstractenLocal spatio-temporal features have been shown to be effective and robust in order to represent simple actions. However, for high level human activities with long-range motion or multiple interactive body parts and persons, the limitation of low-level features blows up because of their localness. This paper addresses the problem by suggesting a framework that computes mid-level features and takes into account their contextual information. First, we represent human activities by a set of mid-level components, referred to as activity components, which have consistent structure and motion in spatial and temporal domain respectively. These activity components are extracted hierarchically from videos, i.e., extracting key-points, grouping them into trajectories and finally clustering trajectories into components. Second, to further exploit the interdependencies of the activity components, we introduce a spatio-temporal context kernel (STCK), which not only captures local properties of features but also considers their spatial and temporal context information. Experiments conducted on two challenging activity recognition datasets show that the proposed approach outperforms standard spatio-temporal features and our STCK context kernel improves further the performance.en
dc.relation.isversionofjnlnamePattern Recognition
dc.relation.isversionofjnlvol45en
dc.relation.isversionofjnlissue12en
dc.relation.isversionofjnldate2012
dc.relation.isversionofjnlpages4182-4191en
dc.relation.isversionofdoihttp://dx.doi.org/10.1016/j.patcog.2012.05.001en
dc.relation.isversionofjnlpublisherElsevieren
dc.subject.ddclabelTraitement du signalen
dc.relation.forthcomingnonen
dc.relation.forthcomingprintnonen


Files in this item

Thumbnail
Name:
yuan-pr2012-action.pdf
Size:
11.25Mb
Format:
PDF
View/Open

This item appears in the following Collection(s)

Show simple item record