Output details
11 - Computer Science and Informatics
Brunel University London
ELVIS: Entertainment-led video summaries
<17> We were the first people to develop a fully-automated technique by combining and analysing real-time physiological responses to generate personalised video summaries applicable across a broad range of content genres. Consequently, significant progress is made in overcoming ‘the semantic gap’ which has challenged the multimedia content modelling community for over two decades. Chenes et al (2013) note the novelty and innovativeness of this approach, which has inspired the Computer Vision and Multimedia Laboratory (CVML) at the University of Geneva to engage in an on-going programme of work to develop automated video highlight detection techniques via user physiological responses.