150 search results - page 4 / 30 » Discriminative Fields for Modeling Semantic Concepts in Vide... |
TASLP
2010
13 years 2 months ago
2010
Abstract--This paper presents a novel method for automatically classifying consumer video clips based on their soundtracks. We use a set of 25 overlapping semantic classes, chosen ...
TOMCCAP
2010
13 years 2 months ago
2010
We investigate the challenging issue of joint audio-visual analysis of generic videos targeting at semantic concept detection. We propose to extract a novel representation, the Sh...
ICMCS
2005
IEEE
14 years 1 months ago
2005
IEEE
For large scale automatic semantic video characterization, it is necessary to learn and model a large number of semantic concepts. A major obstacle to this is the insufficiency o...
ICIP
2007
IEEE
14 years 9 months ago
2007
IEEE
Semantic video indexing is the first step towards automatic video retrieval and personalization. We propose a data-driven stochastic modeling approach to perform both video segmen...
ICASSP
2010
IEEE
13 years 8 months ago
2010
IEEE
Detecting the time of occurrence of an acoustic event (for instance, a cheer) embedded in a longer soundtrack is useful and important for applications such as search and retrieval...