Update README.md
Browse files
README.md
CHANGED
|
@@ -40,3 +40,18 @@ configs:
|
|
| 40 |
- split: train_low
|
| 41 |
path: data/train_low-*
|
| 42 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
- split: train_low
|
| 41 |
path: data/train_low-*
|
| 42 |
---
|
| 43 |
+
# Dataset Card for "beans_gibbons"
|
| 44 |
+
|
| 45 |
+
## Dataset Description
|
| 46 |
+
|
| 47 |
+
**Paper:** https://doi.org/10.1002/rse2.201
|
| 48 |
+
|
| 49 |
+
## Dataset Summary
|
| 50 |
+
|
| 51 |
+
This dataset contains continuous recordings of Hainan gibbon calls with splits and preprocessing like described in BEANS. Many audio files have no calls at all. It is used for **detection** tasks.
|
| 52 |
+
|
| 53 |
+
## Data Splits
|
| 54 |
+
|
| 55 |
+
| train | train_low | valid | test |
|
| 56 |
+
|------|-----| ------ | ------ |
|
| 57 |
+
| 960 | 160 | 480 | 640 |
|