comparison Problems/generateAMT_Learning_Problem.m @ 128:8e660fd14774 ivand_dev

Feature 186
author Ivan Damnjanovic lnx <ivan.damnjanovic@eecs.qmul.ac.uk>
date Mon, 13 Jun 2011 14:55:45 +0100
parents 0211faef9add
children 9c418bea7f6a
comparison
equal deleted inserted replaced
126:db5a7fe1a404 128:8e660fd14774
1 function data = generateAMT_Learning_Problem(nfft, windowSize, overlap) 1 function data = generateAMT_Learning_Problem(nfft, windowSize, overlap)
2 %%% Generate Automatic Music Transcription Problem 2 %% Generate Automatic Music Transcription Problem
3 % 3 %
4 % Centre for Digital Music, Queen Mary, University of London.
5 % This file copyright 2009 Ivan Damnjanovic.
6 %
7 % This program is free software; you can redistribute it and/or
8 % modify it under the terms of the GNU General Public License as
9 % published by the Free Software Foundation; either version 2 of the
10 % License, or (at your option) any later version. See the file
11 % COPYING included with this distribution for more information.
12 %
13 %
14 % generateAMT_Learning_Problem is a part of the SMALLbox and generates 4 % generateAMT_Learning_Problem is a part of the SMALLbox and generates
15 % a problem that can be used for comparison of Dictionary Learning/Sparse 5 % a problem that can be used for comparison of Dictionary Learning/Sparse
16 % Representation techniques in automatic music transcription scenario. 6 % Representation techniques in automatic music transcription scenario.
17 % The function prompts a user for an audio file (mid, wav, mat) reads it 7 % The function prompts a user for an audio file (mid, wav, mat) reads it
18 % and generates a spectrogram given fft size (default nfft=4096), analysis 8 % and generates a spectrogram given fft size (default nfft=4096), analysis
29 % n - number of time points in the spectrogram 19 % n - number of time points in the spectrogram
30 % p - number of dictionary elements to be learned (eg 88 for piano) 20 % p - number of dictionary elements to be learned (eg 88 for piano)
31 % notesOriginal - notes of the original audio to be used for 21 % notesOriginal - notes of the original audio to be used for
32 % comparison (if midi of the original exists) 22 % comparison (if midi of the original exists)
33 % name - name of the audio file to transcribe 23 % name - name of the audio file to transcribe
34 24
25 % Centre for Digital Music, Queen Mary, University of London.
26 % This file copyright 2009 Ivan Damnjanovic.
27 %
28 % This program is free software; you can redistribute it and/or
29 % modify it under the terms of the GNU General Public License as
30 % published by the Free Software Foundation; either version 2 of the
31 % License, or (at your option) any later version. See the file
32 % COPYING included with this distribution for more information.
33 %
35 %% 34 %%
36 FS=filesep; 35 FS=filesep;
37 if ~ exist( 'nfft', 'var' ) || isempty(nfft), nfft = 4096; end 36 if ~ exist( 'nfft', 'var' ) || isempty(nfft), nfft = 4096; end
38 if ~ exist( 'windowSize', 'var' ) || isempty(windowSize), windowSize = 2822; end 37 if ~ exist( 'windowSize', 'var' ) || isempty(windowSize), windowSize = 2822; end
39 if ~ exist( 'overlap', 'var' ) || isempty(overlap), overlap = 0.5; end 38 if ~ exist( 'overlap', 'var' ) || isempty(overlap), overlap = 0.5; end