# speakstream-demo

**Repository Path**: mirrors_apple/speakstream-demo

## Basic Information

- **Project Name**: speakstream-demo
- **Description**: Streaming Text-to-Speech with Interleaved Data
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-08-06
- **Last Updated**: 2026-03-14

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

<div align="center">
  <h1>SpeakStream: Streaming Text-to-Speech with Interleaved Data</h1>

  [![Arxiv](https://img.shields.io/badge/Paper-Arxiv%20Link-green)](https://arxiv.org/pdf/2505.19206)
  [![OpenReview](https://img.shields.io/badge/Paper-OpenReview%20Link-red)](https://openreview.net/forum?id=3x0GD1yvIS)
</div>

In this work, we present a streaming TTS (SpeakStream) that can generate audio incrementally from streaming text using a decoder-only architecture. The model is trained using next-step prediction loss on force-aligned, interleaved text-speech data. During inference SpeakStream generates speech incrementally while absorbing streaming text, making it suitable for cascaded conversational AI agents where an LLM streams text to a TTS system. Our experiments show that SpeakStream matches non-streaming TTS quality while enabling streaming capabilities.

<p align="center">
  <a href="path/to/file.pdf">
    <img src="./assets/speakstream.png" alt="Model Architecture" width="400">
  </a>
  <a href="path/to/file.pdf">
    <img src="./assets/latency.png" alt="Latency results" width="280">
  </a>
</p>

## Generated Samples

The repository provides examples of the generated speech for models trained on [LJSpeech dataset](https://keithito.com/LJ-Speech-Dataset/).

## License

- Repository is released under [LICENSE](LICENSE). 
- All generated speech samples are licensed under [Creative Commons Attribution-Noncommercial-Nonderivatives 4.0 International License](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Citations

```
@article{bai2025speakstream,
  title={SpeakStream: Streaming Text-to-Speech with Interleaved Data},
  author={Bai, He and Gu, Zijin and Likhomanenko, Tatiana and Jaitly, Navdeep},
  journal={arXiv preprint arXiv:2505.19206},
  year={2025}
}
```