Feature Extraction for Audio Fingerprinting Using Wavelet Transform
Active In SP
Joined: Feb 2011
16-02-2011, 10:44 AM
This paper proposes a novel feature extractionscheme for audio fingerprinting using discretewavelet transform (DWT). The proposed schemereduces the granularity, i.e., the minimal length ofaudio, needed for identification in an audiofingerprinting system. The scheme first decomposesthe video frame into sub-bands by DWT
The proposed scheme hastwo advantages:
(1) it needs smaller fingerprintgranularity than other previous work;
(2) it is notonly reliable but also robust against various signal
degradations according to the experimental results
Non-invasive techniques have received much interest in recent years. They are used in lots of applications such as retrieval, recognition and
authentication of digital contents. The non-invasive techniques are performed without modifying the original signal but only analyzing it. Among them, fingerprinting is the most important application that
provides a fast and reliable method for content identification.When the system is presented with an unidentified piece of audio, its fingerprint is extracted and matched against those stored in the database.
Using fingerprints and matching algorithms, distorted versions of a recording can still be identified as the same audio signal   .
This paper is organized as follows. The next section describes some previous work for fingerprint extraction and discusses their granularities
download full report