Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,55 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: torch
|
| 4 |
+
base_model:
|
| 5 |
+
- microsoft/wavlm-large
|
| 6 |
+
pipeline_tag: audio-to-audio
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# ⚡ FocalCodec
|
| 10 |
+
|
| 11 |
+
A low-bitrate single-codebook 16 / 24 kHz speech codec based on [focal modulation](https://arxiv.org/abs/2203.11926).
|
| 12 |
+
|
| 13 |
+
This repository contains the **50 Hz causal checkpoint with a codebook size of 2048** trained on **Libri-Light**, as described in the preprint.
|
| 14 |
+
|
| 15 |
+
- 📜 **Preprint**: https://arxiv.org/abs/2502.04465, https://arxiv.org/abs/1909.xxxxx
|
| 16 |
+
|
| 17 |
+
- 🌐 **Project Page**: https://lucadellalib.github.io/focalcodec-web/
|
| 18 |
+
|
| 19 |
+
- 💾 **GitHub**: https://github.com/lucadellalib/focalcodec
|
| 20 |
+
|
| 21 |
+
<img src="focalcodec-stream.png" width="700">
|
| 22 |
+
|
| 23 |
+
---------------------------------------------------------------------------------------------------------
|
| 24 |
+
|
| 25 |
+
## ▶️ Quickstart
|
| 26 |
+
|
| 27 |
+
See the readme at: https://github.com/lucadellalib/focalcodec
|
| 28 |
+
|
| 29 |
+
---------------------------------------------------------------------------------------------------------
|
| 30 |
+
|
| 31 |
+
## @ Citing
|
| 32 |
+
|
| 33 |
+
```
|
| 34 |
+
@article{dellalibera2025focalcodec,
|
| 35 |
+
title = {{FocalCodec}: Low-Bitrate Speech Coding via Focal Modulation Networks},
|
| 36 |
+
author = {Luca {Della Libera} and Francesco Paissan and Cem Subakan and Mirco Ravanelli},
|
| 37 |
+
journal = {arXiv preprint arXiv:2502.04465},
|
| 38 |
+
year = {2025},
|
| 39 |
+
}
|
| 40 |
+
|
| 41 |
+
@article{dellalibera2025focalcodecstream,
|
| 42 |
+
title = {{FocalCodec-Stream}: Streaming Low-Bitrate Speech Coding via Causal Distillation},
|
| 43 |
+
author = {Luca {Della Libera} and Cem Subakan and Mirco Ravanelli},
|
| 44 |
+
journal = {arXiv preprint arXiv:1909.xxxxx},
|
| 45 |
+
year = {2025},
|
| 46 |
+
}
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
---------------------------------------------------------------------------------------------------------
|
| 50 |
+
|
| 51 |
+
## 📧 Contact
|
| 52 |
+
|
| 53 |
+
[[email protected]](mailto:[email protected])
|
| 54 |
+
|
| 55 |
+
---------------------------------------------------------------------------------------------------------
|