摘要
目前网络数据中包含有大量的多媒体数据,音频作为多媒体信息的重要载体之一,蕴涵了丰富的语义,有效地丰富和补充了人们对信息的感知和获取。而目前的多媒体检索主要利用的是常规的信息检索技术诸如视觉信息,往往忽略了音频信息,这已不能满足需要。音频信息检索就逐渐成为多媒体信息检索的研究重点之一。基于此,针对音频信息检索问题展开讨论,提出一种利用广义模型进行音频分析的新方法。该方法可应用到音频检索中,实验证明了该方法的有效性。
The internet is abundant in multimedia data and audio is one of important information carrier in multimedia. Audio contains abundant semantics, and audio enriches information perception and acquisition. At present, conventional information retrieval methods, such as vision information, are always used in the multimedia retrieval and audio information is ignored. But conventional methods cannot meet the demand. Audio information retrieval is becoming one of research emphases in information retrieval. So the problem of audio retrieval is discussed, and General Model is used to analysis audio in this paper. The method can be used in audio retrieval. Experiments show that the result is well by this method.
出处
《火力与指挥控制》
CSCD
北大核心
2005年第5期52-53,58,共3页
Fire Control & Command Control
基金
江苏省自然科学基金资助项目(BK2004137)
关键词
广义模型
音频分析
语音检索
general model (GM) ,audio analysis ,speech retrieval