Commit
•
a6ba68a
1
Parent(s):
1542755
Update README.md
Browse files
README.md
CHANGED
|
@@ -16,7 +16,7 @@ tags:
|
|
| 16 |
license: cc-by-nc-sa-4.0
|
| 17 |
---
|
| 18 |
|
| 19 |
-
# Model for Age and Gender Recognition based on Wav2vec 2.0
|
| 20 |
|
| 21 |
The model expects a raw audio signal as input and outputs predictions
|
| 22 |
for age in a range of approximately 0...1 (0...100 years)
|
|
@@ -28,6 +28,8 @@ on [aGender](https://paperswithcode.com/dataset/agender),
|
|
| 28 |
[Mozilla Common Voice](https://commonvoice.mozilla.org/),
|
| 29 |
[Timit](https://catalog.ldc.upenn.edu/LDC93s1) and
|
| 30 |
[Voxceleb 2](https://www.robots.ox.ac.uk/~vgg/data/voxceleb/vox2.html).
|
|
|
|
|
|
|
| 31 |
An [ONNX](https://onnx.ai/") export of the model is available from
|
| 32 |
[doi:10.5281/zenodo.7761387](https://doi.org/10.5281/zenodo.7761387).
|
| 33 |
Further details are given in the associated [paper](https://arxiv.org/abs/2306.16962)
|
|
|
|
| 16 |
license: cc-by-nc-sa-4.0
|
| 17 |
---
|
| 18 |
|
| 19 |
+
# Model for Age and Gender Recognition based on Wav2vec 2.0 (6 layers)
|
| 20 |
|
| 21 |
The model expects a raw audio signal as input and outputs predictions
|
| 22 |
for age in a range of approximately 0...1 (0...100 years)
|
|
|
|
| 28 |
[Mozilla Common Voice](https://commonvoice.mozilla.org/),
|
| 29 |
[Timit](https://catalog.ldc.upenn.edu/LDC93s1) and
|
| 30 |
[Voxceleb 2](https://www.robots.ox.ac.uk/~vgg/data/voxceleb/vox2.html).
|
| 31 |
+
In this version of the model we only trained
|
| 32 |
+
the first six transformer layers and.
|
| 33 |
An [ONNX](https://onnx.ai/") export of the model is available from
|
| 34 |
[doi:10.5281/zenodo.7761387](https://doi.org/10.5281/zenodo.7761387).
|
| 35 |
Further details are given in the associated [paper](https://arxiv.org/abs/2306.16962)
|