Datasets:
Search is not available for this dataset
No rows found.
Open cat meows
This dataset is just starting: it holds no meows yet. Contribute with meow contribute after sharing meows in the app.
Meows recorded by Meow Decoder users who chose to share them, each labelled by the person who was with the cat.
Split by cat_id, never at random, or your test set will contain cats your model has already heard.
Better still, also keep each session on one side: meows from one visit share a room and a microphone.
| Column | Meaning |
|---|---|
label |
See below |
cat_id |
Anonymous hash: same cat, same id |
session |
Anonymous hash of one app visit |
device |
phone or computer |
hour |
Local hour of the day, if known |
where |
home, outdoors, vet, car, other, or empty |
contribution_id |
Quote it to have a meow removed |
Labels: F waiting for food, B enjoying attention, I stressed somewhere new, greet saying hello,
out wants to go out, play wants to play, annoyed annoyed, or custom (the owner typed their own words;
the text is only included after review).
- 16 kHz mono wav, CC BY 4.0. Clips with talking in them, or that don't sound like a cat, are not accepted.
- To remove a meow, open an issue at https://github.com/ATHARV-IIT-BOMBAY/meow-decoder with its
contribution_id. - Collected with https://github.com/ATHARV-IIT-BOMBAY/meow-decoder
- Downloads last month
- 65