yangwang825
commited on
Commit
•
aaf6353
1
Parent(s):
b4c7b49
Update README.md
Browse files
README.md
CHANGED
@@ -32,11 +32,11 @@ This system is composed of an ECAPA-TDNN model. It is a combination of convoluti
|
|
32 |
|
33 |
# Performance
|
34 |
|
35 |
-
| Splits | EER(%) | minDCF(0.01) |
|
36 |
-
|
37 |
-
| VoxCeleb1-O | 1.45 | 0.17 |
|
38 |
-
| VoxCeleb1-E | TBD | TBD |
|
39 |
-
| VoxCeleb1-H | TBD | TBD |
|
40 |
|
41 |
# Compute the speaker embeddings
|
42 |
|
|
|
32 |
|
33 |
# Performance
|
34 |
|
35 |
+
| Splits | Backend | S-norm | EER(%) | minDCF(0.01) |
|
36 |
+
|:-------------:|:--------------:|:--------------:|:--------------:|:--------------:|
|
37 |
+
| VoxCeleb1-O | cosine | no | 1.45 | 0.17 |
|
38 |
+
| VoxCeleb1-E | cosine | no | TBD | TBD |
|
39 |
+
| VoxCeleb1-H | cosine | no | TBD | TBD |
|
40 |
|
41 |
# Compute the speaker embeddings
|
42 |
|