DocumentCode :
1984033
Title :
An Improved Onset Detection Algorithm
Author :
Shi Xiang-bin ; Liu Ming ; Wu Jie-hong ; Zhang Lin
Author_Institution :
Sch. of Comput. Sci., Shenyang Univ. of Aeronaut. & Astronaut., Shenyang, China
fYear :
2010
fDate :
20-22 Aug. 2010
Firstpage :
1
Lastpage :
4
Abstract :
Onset detection has an effect on the accuracy of note segmentation and influences the result of Query By Humming(QBH) directly. Based on the Phase Deviation algorithm research, this paper proposes an improved onset detect algorithm called Phase Deviation2(PD2). The special condition where the phases of the adjacent frequency bands change by way of arithmetic sequence is considered in the PD2 algorithm. Therefore, we sum both of the first and second difference of every increases frequency band. Then the average of the sums is used as an onset detection function which is to detect the onset. Experimental results show that the PD2 algorithm has a lower error rate and a higher precision than the PD algorithm on the onset detection.
Keywords :
audio signal processing; music; query processing; signal detection; arithmetic sequence; frequency band; music segmentation; note segmentation; onset detection algorithm; phase deviation algorithm; query by humming; Educational institutions; Electronic mail; Electronics industry; Error analysis; Filtering; Multimedia communication; Music;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Internet Technology and Applications, 2010 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-5142-5
Electronic_ISBN :
978-1-4244-5143-2
Type :
conf
DOI :
10.1109/ITAPP.2010.5566594
Filename :
5566594
Link To Document :
بازگشت