Conclusions
nExplore two approaches to extend the framework of multi-modal news
video retrieval systems
nOverall, our new Text + HLF
retrieval system is able to
nOutperform baseline
system
nAchieve similar results to
top performing automated systems reported in TRECVID 2005
nWhen integrating with other
A/V features, the resulting performance is better than
the best reported result.
nCurrent Work
nInvestigate better ways to link HLFs to queries or general concepts
nBetter correlations between HLFs
nLink to event-based models