BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Topic Models for Human Activity Understanding - Timothy Hospedales
 \, Queen Mary University\, London
DTSTART:20120501T130000Z
DTEND:20120501T140000Z
UID:TALK37789@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Automated modelling of activity in visual surveillance and uns
 tructured consumer multi-media data are important capabilities for securit
 y and commercial content based indexing. These tasks are challenging becau
 se tracking and segmentation may be unreliable in with crowds and occlusio
 n\; manual labelling of sufficient training data may be prohibitively expe
 nsive\; and interesting behaviours may be difficult to model due to being 
 visually subtle relative to background activity and/or defined by complex 
 interactions between multiple objects evolving over time. In this talk\, I
  will describe our work addressing activity modelling with probabilistic t
 opic models in both weakly supervised and unsupervised contexts. Within th
 is framework\, we can address unsupervised or weakly supervised anomaly de
 tection\, hierarchical activity mining and classification\, generalising f
 rom sparse examples and active learning for rare activity discovery.
LOCATION:Small public lecture room\, Microsoft Research Ltd\, 7 J J Thomso
 n Avenue (Off Madingley Road)\, Cambridge
END:VEVENT
END:VCALENDAR