Datasets:
sanchit-gandhi
commited on
Commit
·
d9ae7bd
1
Parent(s):
cdd0de8
more description
Browse files
README.md
CHANGED
|
@@ -47,7 +47,9 @@ task_categories:
|
|
| 47 |
|
| 48 |
The Edinburgh International Accents of English Corpus (EdAcc) is a new automatic speech recognition (ASR) dataset
|
| 49 |
composed of 40 hours of English dyadic conversations between speakers with a diverse set of accents. EdAcc includes a
|
| 50 |
-
wide range of first and second-language varieties of English and a linguistic background profile of each speaker.
|
|
|
|
|
|
|
| 51 |
|
| 52 |
## Supported Tasks and Leaderboards
|
| 53 |
|
|
|
|
| 47 |
|
| 48 |
The Edinburgh International Accents of English Corpus (EdAcc) is a new automatic speech recognition (ASR) dataset
|
| 49 |
composed of 40 hours of English dyadic conversations between speakers with a diverse set of accents. EdAcc includes a
|
| 50 |
+
wide range of first and second-language varieties of English and a linguistic background profile of each speaker.
|
| 51 |
+
Results on latest public, and commercial models show that EdAcc highlights shortcomings of current English ASR models,
|
| 52 |
+
which perform well on existing benchmarks, but degrade significantly on speakers with different accents.
|
| 53 |
|
| 54 |
## Supported Tasks and Leaderboards
|
| 55 |
|