root
commited on
Commit
·
29f8ef1
1
Parent(s):
62e3071
update
Browse files
README.md
CHANGED
|
@@ -59,9 +59,12 @@ This dataset is meticulously curated from five renowned open-source emotional di
|
|
| 59 |
| Total Duration | ~51 hours |
|
| 60 |
| Languages | Chinese, English |
|
| 61 |
| Language Ratio (CN:EN) | Approx. 1:1 |
|
| 62 |
-
| Sampling Rate | 16kHz
|
| 63 |
| Avg. Expressiveness Score (DeEAR) | 80.2 |
|
| 64 |
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
|
| 67 |
|
|
|
|
| 59 |
| Total Duration | ~51 hours |
|
| 60 |
| Languages | Chinese, English |
|
| 61 |
| Language Ratio (CN:EN) | Approx. 1:1 |
|
| 62 |
+
| Sampling Rate | 16kHz |
|
| 63 |
| Avg. Expressiveness Score (DeEAR) | 80.2 |
|
| 64 |
|
| 65 |
+
## Our Screening Tool: DeEAR
|
| 66 |
+
|
| 67 |
+
The high expressiveness of this dataset was achieved using our screening tool, **DeEAR**. If you need to build larger batches of high-expressiveness data yourself, you are welcome to use this tool. You can find it on our [GitHub](https://github.com/FreedomIntelligence/ExpressiveSpeech).
|
| 68 |
|
| 69 |
|
| 70 |
|