Quickstart¶
The most common task is to load a database
with audb.load().
Let’s first see which databases are available on our public Artifactory server.
import audb
audb.available(only_latest=True)
| backend | host | repository | version | |
|---|---|---|---|---|
| name | ||||
| air | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.4.2 |
| cough-speech-sneeze | artifactory | https://audeering.jfrog.io/artifactory | data-public | 2.0.1 |
| crema-d | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.2.0 |
| emodb | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.4.1 |
| micirp | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.0.0 |
| musan | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.0.0 |
| vadtoolkit | artifactory | https://audeering.jfrog.io/artifactory | data-public | 1.1.0 |
Let’s load version 1.4.1 of the emodb database.
db = audb.load("emodb", version="1.4.1", verbose=False)
This downloads the database header,
all the media files,
and tables with annotations
to a caching folder on your machine.
The database is then returned
as an audformat.Database object.
Each database comes with a description, which is a good starting point to learn what the database is all about.
db.description
'Berlin Database of Emotional Speech. A German database of emotional utterances spoken by actors recorded as a part of the DFG funded research project SE462/3-1 in 1997 and 1999. Recordings took place in the anechoic chamber of the Technical University Berlin, department of Technical Acoustics. It contains about 500 utterances from ten different actors expressing basic six emotions and neutral.'
The annotations of a database are stored in
tables represented by audformat.Table.
db.tables
emotion:
type: filewise
columns:
emotion: {scheme_id: emotion, rater_id: gold}
emotion.confidence: {scheme_id: confidence, rater_id: gold}
emotion.categories.test.gold_standard:
type: filewise
split_id: test
columns:
emotion: {scheme_id: emotion, rater_id: gold}
emotion.confidence: {scheme_id: confidence, rater_id: gold}
emotion.categories.train.gold_standard:
type: filewise
split_id: train
columns:
emotion: {scheme_id: emotion, rater_id: gold}
emotion.confidence: {scheme_id: confidence, rater_id: gold}
files:
type: filewise
columns:
duration: {scheme_id: duration}
speaker: {scheme_id: speaker}
transcription: {scheme_id: transcription}
Each table contains columns (audformat.Column)
that have corresponding schemes (audformat.Scheme)
describing its content.
For example,
to get an idea about the emotion annotations
stored in the emotion column,
we can inspect the corresponding scheme.
db.schemes["emotion"]
description: Six basic emotions and neutral.
dtype: str
labels: [anger, boredom, disgust, fear, happiness, sadness, neutral]
Finally, we get the actual annotations
as a pandas.DataFrame.
df = db["emotion"].get() # get table
df[:3] # show first three entries
| emotion | emotion.confidence | |
|---|---|---|
| file | ||
| /home/runner/audb/emodb/1.4.1/d3b62a9b/wav/03a01Fa.wav | happiness | 0.90 |
| /home/runner/audb/emodb/1.4.1/d3b62a9b/wav/03a01Nc.wav | neutral | 1.00 |
| /home/runner/audb/emodb/1.4.1/d3b62a9b/wav/03a01Wa.wav | anger | 0.95 |