|
|
--- |
|
|
language: en |
|
|
tags: |
|
|
- audio |
|
|
- speech |
|
|
- target-speaker-extraction |
|
|
- source-separation |
|
|
- speakerbeam |
|
|
- dprnn |
|
|
license: apache-2.0 |
|
|
datasets: |
|
|
- wsj0-2mix |
|
|
- librimix |
|
|
- wham |
|
|
- whamr |
|
|
library_name: pytorch |
|
|
--- |
|
|
|
|
|
# TSExcalibur |
|
|
|
|
|
🎯 **TSExcalibur** is a modular and extensible toolkit for **Target Speaker Extraction (TSE)**. |
|
|
|
|
|
> **GitHub Repository**: [youzhenghai/TSExcalibur](https://github.com/youzhenghai/TSExcalibur) |
|
|
|
|
|
--- |
|
|
|
|
|
## 🧠 Features |
|
|
|
|
|
|
|
|
- Implementations of leading TSE models |
|
|
|
|
|
|
|
|
--- |
|
|
|
|
|
|
|
|
## 🚀 Usage |
|
|
|
|
|
Coming soon! In the meantime, you can clone the repo and run: |
|
|
|
|
|
```bash |
|
|
git clone https://github.com/youzhenghai/TSExcalibur.git |
|
|
cd TSExcalibur |
|
|
# Set up your environment and run training or inference scripts |
|
|
|
