Fixed distortion segmentation in efficient sound segment searching

Searching query signal from stored signal is formulated as a segment searching problem where signal is converted into a sequence of feature vectors. As an efficient segment searching algorithm, a new pruning method in the segment sequence has been proposed and the effectiveness has been shown through experimental evaluation. The proposed searching algorithm is 20 30 times faster than the conventional Active Searching algorithm. As the first step of the proposed method distortion based segmentation is carried out. As searching criterion is based on norm, the segmentation is expected to be carried out based on criterion. This paper compares three segmentation methods; maximum distortion segmentation, average distortion segmentation and fixed length segmentation. The average distortion and fixed length segmentation methods are very efficient. On the other hand, the maximum distortion segmentation produces fixed distortion segmentation and does not require radius information. The experimental results show that first two methods have almost equal performance in segment searching when the number of segments are the same.