a neural network toolbox for animal vocalizations and bioacoustics#
vak is a library that makes it easier
for researchers studying animal
vocalizations—such as birdsong, bat calls,
and even human speech—to work with
neural network algorithms.
To learn more about the goals and design of vak, please see this talk from the SciPy 2023 conference, and the associated Proceedings paper here.
Currently, the main use is automated annotation of animal vocalizations. By annotation, we mean something like this example of annotated birdsong:
Please see links below for information on how to get started and how to use
apply neural network models to your data.
If you are new to working with
and you’re looking for installation instructions and a tutorial,
If there is something specific you’re trying to do,
like use your own spectrogram files or annotation formats with
please check in the How-to guides.
For help, please begin by checking out the Frequently Asked Questions (FAQs).
To ask a question about vak, discuss its development,
or share how you are using it,
please start a new “Q&A” topic on the VocalPy forum
with the vak tag:
To report a bug, or to request a feature,
please use the issue tracker on GitHub:
If you need to look up information about the command-line interface, configuration files, etc., please consult the Reference.
To learn about development of
vak and how you can contribute, please see Development
For more about the goals of
vak and its development, please see About vak.
Not enough open-source research software libraries have poems. We do, here: Poems.