期刊文献+

基于音频子带能量动态范围的快速音频检索

Audio Retrieval Based on Dynamic Range of Sub-band Energy Feature
下载PDF
导出
摘要 提出了利用音频子带能量动态范围特征实现两阶段快速音频检索的方法。在预处理阶段根据音频库的子带能量动态范围(DRSBE)特征首先建立1个索引库,检索时分为2步:第一步先计算输入参考音频片段的DRSBE特征,然后根据数据库中建立的索引找到候选音频;第二步计算参考音频和候选音频之间的相似度,输出最后结果。实验结果表明,基于DRSBE特征的快速音频检索方法对于同源音频检索的速度和精度都非常高,在高质量的广播音频检索中达到了实用要求。 A two-stage quick audio retrieval method is proposed based on the DRSBE (Dynamic Range of Sub- Band Energy) feature. In the perprocessing, an audio indexing database is first constructed according to the DRSBE feature of the audio in the database. The audio retrieval method is carried out in two stages. Firstly, the DRSBE feature of the input reference audio clip is calculated, and then the audio clip candidates are quickly extracted from the database by the indexing method. Secondly, the accurate similarity between the reference audio and the audio candidates is evaluated to refine the final output. The experiment show that this quick audio retrieval method based on DRSBE feature has good performance in speed and precision and is good enough for the retrieval of high quality audio in broadcasting application.
出处 《电声技术》 2009年第10期66-68,72,共4页 Audio Engineering
关键词 音频检索 子带能量 动态范围 索引 audio retrieval sub-band energy dynamic range index
  • 相关文献

参考文献9

  • 1WANG Y,LIU Z,HUANG J C. Multimedia content analysisusing both audio and visual clues[J]. IEEE Signal Processing Magazine,2000,17(6) : 12-36.
  • 2ZHANG Wei-qiang, LIU Jia. Two-stage method for specific audio retrieval [C]// Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing. Honolulu :IEEE Press , 2007 , Ⅳ : 85-88.
  • 3SMITH G, MURASE H, KASHINO K. Quick audio retrieval using active search [C]// Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing. Seattle: IEEE Press, 1998 : 3777-3780.
  • 4KASHINO K, KUROZUMI T, MURASE H. A quick search method for audio and video signals based on histogram pruning[J]. IEEE Trans. on Multimedia, 2003, 5(3) :348-357.
  • 5DUDA R,HART P,STOCK D. Pattern classification[M]. [S.l.]:John Wiley & Sons,2000.
  • 6KIM K M,KIM S Y,JEON J K, et al. Quick audio retrieval using multiple feature vectors[J]. IEEE Trans. on Consumer Electronics, 2006,52 ( 1 ) : 200-205.
  • 7LOGAN B. Mel frequency cepstral coefficients for music modeling [C]// Proceeding of the Internatioal Symposium on Music Information Retrieval ( ISMIR ), 2000.
  • 8蔡莲红,黄德智,蔡锐.现代语音基础与应用[M].北京:清华大学出版社,2003.
  • 9CHENG D Y,Gersho A,Ramamurthi B,et al. Fast search algorithm for vector quantization and pattern matching[C]// Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing. San Diego: IEEE Press, 1984,9 : 372-375.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部