9
$\begingroup$

What I found so far was either bird datasets (xeno-canto or all except the bees from the bioacoustics dataset page) or bees (also on the bioacoustics dataset page). I know there is the tierstimmenarchiv by the Museum für Naturkunde Berlin, but this is not downloadable as a whole. I am not looking for curated or labeled datasets, just for a diverse set of audio files.

I also looked through kaggle without much success.

Improve this question
$\endgroup$
1
  • $\begingroup$ Hi Julian. Are marine sound datasets of interest for your search, or not? $\endgroup$
    – Dan Stowell ♦
    Commented Aug 24, 2022 at 13:59

4 Answers 4

Reset to default
7
$\begingroup$
  • The webpage "Bioacoustics Datasets" (new in 2024!) now contains a big listing of many datasets, including lots of non-bird as well as bird audio. Mammals, insects, anurans and more.
  • If what you want is diversity more than "quality" (e.g. taxonomic precision), you can look at AudioSet (a generic sound dataset from Youtube). For example the AudioSet "animal" tag contains over 40,000 annotated sound clips, though approx half of those are cats and dogs.
  • There are large marine sound datasets, e.g. this DCLDE dataset or DOCC10
  • bio.acousti.ca contains diverse animal sound recordings, across many taxa, and taxonomically organised. It certainly covers mammal and insect sounds, for example. I believe the data is downloadable in bulk, though I'm not 100% sure of the process.
  • Through GBIF you can search for occurrences with audio media. To give an arbitrary example, a GBIF search for Surica suricatta currently shows 4 results, all via Tierstimmenarchiv. It should, I think, be possible to use GBIF's API to automatically download files.

I look forward to reading other people's suggestions of data sources too!

Improve this answer
$\endgroup$
1
  • $\begingroup$ Perfect, thanks, Dan! $\endgroup$
    – Julian Zimmermann
    Commented Aug 24, 2022 at 18:45
3
$\begingroup$

LILA BC is a website that hosts biological datasets with the goal of making them easier to find and access for machine learning researchers/practitioners. Right now it's entirely image datasets, but they have a list of other bioacoustic datasets. Looks like it is mostly marine mammals and birds.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Cool project, thanks! $\endgroup$
    – Julian Zimmermann
    Commented Aug 24, 2022 at 18:47
1
$\begingroup$

Macaulay Library has audio recordings of hundreds of non-avian species that researchers can request free of charge. Archives include, for example:

Improve this answer
$\endgroup$
0
$\begingroup$

You might want to check dosits.org for a good database of underwater sounds from a good variety of taxa.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Hi Dennis, thanks for your answer and the link. Do you know if it is possible to download the whole archive? I was unable to find a way to download the sounds other than scrapping it. $\endgroup$
    – Julian Zimmermann
    Commented Aug 27, 2022 at 8:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.