UEC at TRECVID 2007 High Level Feature Task

In this paper, we describe our approach and results for high-level feature extraction task at TRECVID 2007. This year, we adopted late fusion of several types of features. As a first step, we extract several types of visual features and ASR texts from the given movies, and apply SVM to them independently. As the next step, we fuse these results by linear combination with weights chosen by cross validation.

• Type: Conference
• Year: 2007
• Where: TRECVID
• Authors: O. Liu, Z. Tang, K. Yanai
• Keywords: ASR Texts, Late Fusion, TRECVID 2007, TRECVID 2008, first Step