EASIER Sampling for Audio Event Identification
暂无分享,去创建一个
An audio event refers to some specific audio sound which plays important role for video content analysis. In our previous work [M. Xu et. al., (2004)], we have established audio event identification as an audio classification task. Due to the large size of audio database, representative samples are necessary for training the classifier. However, the commonly used random selection of training samples is often not adequate in selecting representative samples. In this paper we present EASIER sampling algorithm to select those data which more efficiently represent audio data characters for audio event identifier training. EASIER deterministic ally produces a subsample whose "distance" from the complete database is minimal. Experiments in the context of audio event identification show that EASIER outperforms simple random sampling significantly
[1] Bin Chen,et al. Efficient data reduction with EASE , 2003, KDD '03.
[2] Qi Tian,et al. HMM-Based Audio Keyword Generation , 2004, PCM.
[3] Anoop Gupta,et al. Automatically extracting highlights for TV Baseball programs , 2000, ACM Multimedia.
[4] Changsheng Xu,et al. Audio keyword generation for sports video analysis , 2004, MULTIMEDIA '04.
[5] Surya Nepal,et al. Automatic detection of 'Goal' segments in basketball videos , 2001, MULTIMEDIA '01.