English
资源说明：Fast Music Indexing PoC based on Shazam
freezam 0.4 ----------- This is an implementation of the fast audio identification algorithm designed by Shazam. We use an approximate technique, but only use FOSS libraries and tools to get the work done. What you should know is that this is a mere PoC developed at Mark Hensen's stats class at UCLA. If you want the secret sauce, talk to the folks at Shazam. ** This project has been discontinued. It was kept here only for historical and entertainment purposes. ** How does it work ---------------- Shazam is composed by a set of tools for spectrogram analysis, fingerprinting, storage, and fast retrieval. Here we do not deal with the storage and retrieval part, only with the preparation of the files and the technique of identification. The file 'PoC_shazam.R' is commented and is self-explanatory. There are a few steps to take: 0. Transform music spectrograms into matrices 1. Analyze matrices, storing its 'fingerprint' and generating unique hash numbers 2. Use stored fingerprints and use them for music identification Requirements ------------ Before you install, make sure you have met the dependencies: - Gstreamer 1.0 and its plugins (for the 'good' set) - R-base - "Sound" and "signal" packages for R License ------- freezam is licensed under GPL, version 2. Read LICENSE file for details. Authors ------- Read AUTHORS file. Thanks ------ Mark Hansen at UCLA for showing us the way and making stats awesome.
本源码包内暂不包含可直接显示的源代码文件，请下载源码包。
