Update README.md
Browse files
README.md
CHANGED
|
@@ -2,9 +2,12 @@
|
|
| 2 |
license: mit
|
| 3 |
---
|
| 4 |
ABX-accent
|
|
|
|
|
|
|
| 5 |
The ABX-accent project is based on the preparation and evaluation of the Accented English Speech Recognition Challenge (AESRC) dataset [1], using ABXpy for evaluation [2][3]. This repository provides all the necessary tools and resources to carry out both dataset preparation and evaluation.
|
| 6 |
|
| 7 |
What is ABX Evaluation?
|
|
|
|
| 8 |
The ABXpy metric evaluates whether a representation X of a speech unit (e.g., the word “bap”) is closer to a same-category example A (also “bap”) than to a different-category example B (e.g., “bop”). The ABX error rate is calculated by averaging the classification errors over all minimal phoneme trigrams in the corpus.
|
| 9 |
|
| 10 |
This benchmark focuses on the more challenging ABX across-speaker task, where the X example is spoken by a different speaker than the ones in pair (A, B), testing speaker-invariant phonetic discrimination.
|
|
|
|
| 2 |
license: mit
|
| 3 |
---
|
| 4 |
ABX-accent
|
| 5 |
+
-----------
|
| 6 |
+
|
| 7 |
The ABX-accent project is based on the preparation and evaluation of the Accented English Speech Recognition Challenge (AESRC) dataset [1], using ABXpy for evaluation [2][3]. This repository provides all the necessary tools and resources to carry out both dataset preparation and evaluation.
|
| 8 |
|
| 9 |
What is ABX Evaluation?
|
| 10 |
+
-----------------------
|
| 11 |
The ABXpy metric evaluates whether a representation X of a speech unit (e.g., the word “bap”) is closer to a same-category example A (also “bap”) than to a different-category example B (e.g., “bop”). The ABX error rate is calculated by averaging the classification errors over all minimal phoneme trigrams in the corpus.
|
| 12 |
|
| 13 |
This benchmark focuses on the more challenging ABX across-speaker task, where the X example is spoken by a different speaker than the ones in pair (A, B), testing speaker-invariant phonetic discrimination.
|