Update README.md
Browse files
README.md
CHANGED
|
@@ -1,7 +1,18 @@
|
|
| 1 |
---
|
| 2 |
license: cc-by-sa-4.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
| 4 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
# Usage
|
| 6 |
|
| 7 |
``` sh
|
|
|
|
| 1 |
---
|
| 2 |
license: cc-by-sa-4.0
|
| 3 |
+
tags:
|
| 4 |
+
- speaker extraction
|
| 5 |
+
- speech separation
|
| 6 |
+
- lip
|
| 7 |
+
- face
|
| 8 |
+
- multimodal
|
| 9 |
+
- cocktail party problem
|
| 10 |
---
|
| 11 |
|
| 12 |
+
A modified version of the VoxCeleb2 Dataset. Original data can be downloaded [here](https://www.robots.ox.ac.uk/~vgg/data/voxceleb/vox2.html).
|
| 13 |
+
|
| 14 |
+
This dataset is used for Audio-visual speaker extraction conditioned on face recordings in [the reentry paper](https://ieeexplore.ieee.org/document/9721129), which the code can be found [here](https://github.com/zexupan/reentry).
|
| 15 |
+
|
| 16 |
# Usage
|
| 17 |
|
| 18 |
``` sh
|