knn

stanford_mir/core.py

 import errno
+import librosa
+import numpy
 import os
 import os.path
+import sklearn
 import urllib
 
-def download_drum_samples(path='drum_samples'):
+def extract_features(signal, features):
+    fvs = list()
+    for feature_name in features:
+        if feature_name == 'zero_crossing_rate':
+            fvs.append( librosa.feature.zero_crossing_rate(signal)[0, 0] )
+        elif feature_name == 'spectral_centroid':
+            fvs.append( librosa.feature.spectral_centroid(signal)[0, 0] )
+    return fvs
+
+def get_features(collection='drum_samples_train',
+                 features=('zero_crossing_rate', 'spectral_centroid'),
+                 scaler=None):
+    if collection == 'drum_samples_train':
+        kick_filepaths, snare_filepaths = download_samples('drum_samples_train')
+        kick_signals = [
+            librosa.load(p)[0] for p in kick_filepaths
+        ]
+        snare_signals = [
+            librosa.load(p)[0] for p in snare_filepaths
+        ]
+
+        kick_features = numpy.array([extract_features(x, features) for x in kick_signals])
+        snare_features = numpy.array([extract_features(x, features) for x in snare_signals])
+        feature_table = numpy.vstack((kick_features, snare_features))
+        if scaler is None:
+            scaler = sklearn.preprocessing.MinMaxScaler(feature_range=(-1, 1))
+            scaler.fit(feature_table)
+        training_features = scaler.transform(feature_table)
+
+        kick_labels = numpy.zeros(10)
+        snare_labels = numpy.ones(10)
+        training_labels = numpy.concatenate((kick_labels, snare_labels))
+
+        return training_features, training_labels, scaler
+
+    elif collection == 'drum_samples_test':
+        kick_filepaths, snare_filepaths = download_samples('drum_samples_test')
+        kick_signals = [
+            librosa.load(p)[0] for p in kick_filepaths
+        ]
+        snare_signals = [
+            librosa.load(p)[0] for p in snare_filepaths
+        ]
+
+        kick_features = numpy.array([extract_features(x, features) for x in kick_signals])
+        snare_features = numpy.array([extract_features(x, features) for x in snare_signals])
+        feature_table = numpy.vstack((kick_features, snare_features))
+        if scaler is None:
+            scaler = sklearn.preprocessing.MinMaxScaler(feature_range=(-1, 1))
+            scaler.fit(feature_table)
+        test_features = scaler.transform(feature_table)
+
+        kick_labels = numpy.zeros(30)
+        snare_labels = numpy.ones(30)
+        labels = numpy.concatenate((kick_labels, snare_labels))
+
+        return test_features, labels, scaler
+
+def download_samples(collection='drum_samples_train'):
     """Download ten kick drum samples and ten snare drum samples.
 
-        `path`: output directory containing the twenty drum samples
+        `collection`: output directory containing the twenty drum samples
 
     Returns:
 
@@ -15,17 +76,27 @@ def download_drum_samples(path='drum_samples'):
         `snare_filepaths`: list of snare drum filepaths
     """
     try:
-        os.makedirs(path)
+        os.makedirs(collection)
     except OSError as exception:
         if exception.errno != errno.EEXIST:
             raise
         else:
-            print "Directory %s already exists." % path
+            print "Directory %s already exists." % collection
+    if collection == 'drum_samples_train':
         for drum_type in ['kick', 'snare']:
             for i in range(1, 11):
                 filename = '%s_%02d.wav' % (drum_type, i)
                 urllib.urlretrieve('http://audio.musicinformationretrieval.com/drum_samples/%s' % filename,
-                               filename=os.path.join(path, filename))
-    kick_filepaths = [os.path.join(path, 'kick_%02d.wav' % i) for i in range(1, 11)]
-    snare_filepaths = [os.path.join(path, 'snare_%02d.wav' % i) for i in range(1, 11)]
+                                   filename=os.path.join(collection, filename))
+        kick_filepaths = [os.path.join(collection, 'kick_%02d.wav' % i) for i in range(1, 11)]
+        snare_filepaths = [os.path.join(collection, 'snare_%02d.wav' % i) for i in range(1, 11)]
+        return kick_filepaths, snare_filepaths
+    elif collection == 'drum_samples_test':
+        for drum_type in ['kick', 'snare']:
+            for i in range(30):
+                filename = '%s_%02d.wav' % (drum_type, i)
+                urllib.urlretrieve('http://audio.musicinformationretrieval.com/drum_samples/test/%s' % filename,
+                                   filename=os.path.join(collection, filename))
+        kick_filepaths = [os.path.join(collection, 'kick_%02d.wav' % i) for i in range(30)]
+        snare_filepaths = [os.path.join(collection, 'snare_%02d.wav' % i) for i in range(30)]
         return kick_filepaths, snare_filepaths