Quickstart

Browse available datasets and select one. We load emodb, which is returned as an audformat.Database.

>>> db = audb.load("emodb", version="1.4.1")

Inspect label schemes, and request emotion labels for the test split.

>>> list(db.schemes)
['age',
 'confidence',
 'duration',
 'emotion',
 'gender',
 'language',
 'speaker',
 'transcription']
>>> db.get("emotion", splits="test")  # returns dataframe
                       emotion
file
.../wav/12a01Fb.wav  happiness
.../wav/12a01Lb.wav    boredom
.../wav/12a01Nb.wav    neutral
.../wav/12a01Wc.wav      anger
.../wav/12a02Ac.wav       fear
...                        ...

Or inspect tables, and request labels from a table.

>>> list(db)
['emotion',
 'emotion.categories.test.gold_standard',
 'emotion.categories.train.gold_standard',
 'files',
 'speaker']
>>> db["emotion.categories.test.gold_standard"].get()  # returns dataframe
                       emotion  emotion.confidence
file
.../wav/12a01Fb.wav  happiness                0.95
.../wav/12a01Lb.wav    boredom                0.90
.../wav/12a01Nb.wav    neutral                0.95
.../wav/12a01Wc.wav      anger                0.95
.../wav/12a02Ac.wav       fear                0.90
...                        ...                 ...

Load a media file, selected from the index of the dataframe or from the files index db.files.

>>> import audiofile
>>> signal, sampling_rate = audiofile.read(db.files[0])

Listen to the signal.

>>> import sounddevice
>>> sounddevice.play(signal.T, sampling_rate)