Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,15 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pretty_name: RS-C
|
| 6 |
+
size_categories:
|
| 7 |
+
- 100K<n<1M
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
ReefSet-C (RS-C) serves as a benchmark for [DHAuDS](https://arxiv.org/abs/2511.18421) in the context of test-time adaptation for audio classification.
|
| 11 |
+
RS-C, a subset of [ReefSet](https://arxiv.org/abs/2404.16436), utilizes only test-set samples to address domain shift. Each sample in RS-C has a duration
|
| 12 |
+
of 1.88 seconds, a sample rate of 16 kHz, and belongs to one of 37 classes. The dataset comprises seven corruption
|
| 13 |
+
sets: WHN, ENQ, END1, END2, ENSC, TST, and PSH. Each corruption set contains two levels, L1 and L2, with L2 representing
|
| 14 |
+
a higher degree of complexity than L1. RS-C includes a total of 239,918 samples, and each corruption level, such as
|
| 15 |
+
WHN-L1 or WHN-L2, consists of 17,137 samples.
|