Tag Archives: wav2vec

Predict emotional states with the audEERING model

audEERING recently published an emotion prediction model based on a finetuned Wav2vec2 transformer model.

Here I'd like to show you how you can use this model to predict your audio samples (it is actually also explained in the Github link above).

As usual, you should start with dedicating a folder on your harddisk for this and install a virtual environment:

virtualenv -p=3 venv

which means we want python version 3 (and not 2)
Don't forget to activate it!

Then you would need to install the packages that are used:

protobuf == 3.20

easiest to copy this list into a file called requierments.txt and then do

pip install -r requirements.txt

and start writing a python script that includes the packages:

import audeer
import audonnx
import numpy as np
import audiofile
import audinterface

, load the model:

# and download and load the model
url = 'https://zenodo.org/record/6221127/files/w2v2-L-robust-12.6bc4a7fd-1.1.0.zip'
cache_root = audeer.mkdir('cache')
model_root = audeer.mkdir('model')

archive_path = audeer.download_url(url, cache_root, verbose=True)
audeer.extract_archive(archive_path, model_root)
model = audonnx.load(model_root)

sampling_rate = 16000
signal = np.random.normal(size=sampling_rate).astype(np.float32)

load a test sentence (in 16kHz 16 bit wav format)

# read in a wave file for testing
signal, sampling_rate = audiofile.read('test.wav')

and print out the results

# print the results in the order arousal, dominance, valence.
print(model(signal, sampling_rate)['logits'].flatten())

You can also use audinterace's magic and process a whole list of files like this:

# define the interface
interface = audinterface.Feature(
        'outputs': 'logits',
# create a list of audio files
files = ['test.wav']
# and process it

should result in:

Also check out this great jupyter notebook from audEERING

How to set up wav2vec embedding for nkululeko

Since version 0.10, nkululeko supports facebook's wav2vec 2.0 embeddings as acoustic features.
This post shows you how to set this up.

set up nkululeko

in your nkululeko configuration (*.ini) file, set the feature extractor as wav2vec2 and denote the path to the model like this:

type = ['wav2vec2']
wav2vec.model = /my path to the huggingface model/

Alternatively you can state the huggingface model name directly:

type = ['wav2vec2-base-960h']

Out of the box, as embeddings the last hidden layer is used. But the original wav2vec2 model consists of 7 CNN layers followed by up to 24 transformer layers. if you like to use an earlier layer than the last one, you can simply count down-

type = wav2vec2
wav2vec.layer = 12

This would use the 12th layer of a 24 layer model and only the4 CNN layers of a 12 layer model.