Learning to Recognise Dynamic
Visual Content from Broadcast Footage is a 4 year project
funded by the EPSRC
which brings together the Centre
for Vision Speech and Signal Processing at the University
of Surrey, the Visual
Geometry Group at the University
of Oxford and the Computer
Vision Group at the University
of Leeds to tackle the subject of automatically learn to
recognise dynamic activity in broadcast footage with
demonstration activities in both Sign Language and more
general actions and activity. The task is to use linguistic
annotation provided by subtitle text and scripts as weak
supervision in the learning process. The project started in
late 2011 and will run until 2015. As we progress, this site
will be updated with results and
publications so check back
soon. For more details contact the lead academics at each
site.
