2026-07-28T21:37:06Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/4925 2021-12-13T06:22:34Z 1582963302567:1597824273898 user-ucsy

Audio Fingerprinting based on Wavelet Spectral Entropy Linn, Kyi Chan Nyein Audio-Fingerprints (AFPs) are essentialcharacteristics of digital audio streams used to scorethe perceptual similarity between audio signals.Audio-fingerprinting systems extract features fromthe signal normally on a frame by frame basis. Inthis paper, a robust audio-fingerprint (AFP)approach is developed based on spectral entropy inwavelet domain. To extract the fingerprints of asong, Shannon’s entropy is determined from thespectral coefficients of each one of the first 24critical bands according to the Bark scale. Theperformance of this AFP system is evaluated on amusic database containing various genres. Therobustness of the system is validated throughdegraded music signals in 4 different ways: whitenoise addition, lossy compression, lowpass filter andresampling. 2010-12-16 http://hdl.handle.net/20.500.12678/0000004925 https://meral.edu.mm/records/4925