1 Hit in 4.1 sec

TT+GT at TRECVID 2010 Workshop

Nakamasa Inoue, Toshiya Wada, Yusuke Kamishima, Koichi Shinoda, Ilseo Kim, Byungki Byun, Chin-Hui Lee
2010 TREC Video Retrieval Evaluation  
Initially, we hoped F A TT+GT+run3 3 and F A TT+GT+run4 4 would return at least similar results, but the kernelized MFoM approach did not provide good results compared to the run from Tokyo Tech side,  ...  F A TT+GT run4 4 For the model, F A TT+GT+run4 4, the kernelized MFoM approach described in Section 2.2.3 was used.  ...  We predict position p c (i + 1) of the same cluster at the next frame as Next we associate this cluster c with the cluster nearest to p c (i + 1).  ... 
dblp:conf/trecvid/InoueWKSKBL10 fatcat:wcfbktyz5fb6xo5kz5kjwo25ku