Berlin EmoDB

A database of emotional speech. Ten actors (5 female and 5 male) simulated the emotions,
producing 10 German utterances (5 short and 5 longer sentences) which could be used in everyday communication
and are interpretable in all applied emotions.
The recordings were taken in an anechoic chamber with high-quality recording equipment. In addition to the sound
electro-glottograms were recorded. The speech material comprises about 800 sentences (seven emotions * ten actors * ten sentences + some second versions).
The complete database was evaluated in a perception test regarding the recognisability of emotions and their
naturalness. Utterances recognised better than 80% and judged as natural by more than 60% of the listeners were included in the database, resulting in 535 samples.

They are split into 4 speakers in test and 6 speakers in train.

The database meta information is in audformat.

The audio format is 16kHz wav format.

Here's a sample script in Python how to get train and test dataframes:

import audformat

!wget https://zenodo.org/record/7447302/files/emodb.zip

!unzip emodb.zip

db = audformat.Database.load('emodb')

df_test = db['emotion.categories.test.gold_standard'].get()

df_train = db['emotion.categories.train.gold_standard'].get()

print(f'samples in train: {df_train.shape[0]}, and test: {df_test.shape[0]}')