|
Title:
|
RETRIEVAL OF AUDIO CLIPS REPRESENTED BY MINIMUM SPANNING TREE |
|
Author(s):
|
Lin Jie , Zhao Xue-yan , Yang Jian-gang |
|
ISBN:
|
972-98947-3-6 |
|
Editors:
|
Nuno Guimarães and Pedro Isaías |
|
Year:
|
2004 |
|
Edition:
|
Single |
|
Keywords:
|
Audio Retrieval, Minimum Spanning Tree, Audio Relevance Feedback. |
|
Type:
|
Full Paper |
|
First Page:
|
1067 |
|
Last Page:
|
1072 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
This paper introduces a new and efficient audio retrieval algorithm by unsupervised learning which is different with prior methods in audio clip retrieval that needs to generate audio template for audio database by supervised learning and find similar audio clip based on pre-trained audio templates. The procedure of this new algorithm: first, extracting features directly from compressed domain; second, Generating limited number of centroids through the clustering of Minimum Spanning Tree (MST), and the clustering centroids are used to represent each audio clip and performed efficient matching of audio clips; finally, in order to guarantee retrieval results consistent with users perception, the update of feature weights and centroids are both used in this paper. The experiment shows that the audio retrieval based on MST is robust against noise. |
|
|
|
|
|
|
