Metadata-Version: 2.4
Name: voice-detector
Version: 0.0.1
Summary: Name reservation stub - install from pip.muid.io for full package
Project-URL: Homepage, https://pip.muid.io/simple/voice-detector/
Project-URL: Documentation, https://pip.muid.io/simple/voice-detector/
License: MIT
Keywords: audio,detection,diarization,speaker,voice
Classifier: Development Status :: 1 - Planning
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Python: >=3.9
Description-Content-Type: text/markdown

# voice-detector

**This is a name reservation stub.** The full package is hosted on a private PyPI server.

## Installation

To install the full `voice-detector` package with all features:

```bash
pip install voice-detector --extra-index-url https://pip.muid.io/simple/
```

## What is voice-detector?

Multi-pass voice/speaker switch detection SDK with CLI. Features:

- **Silero VAD** - Voice Activity Detection
- **Parselmouth** - Pitch analysis (Praat)
- **Resemblyzer** - Speaker embeddings (d-vectors)
- **SpeechBrain** - Optional GPU-accelerated backend

### CLI Usage

```bash
vdetect analyze audio.mp3 --preset podcast --output result.json
```

### Python API

```python
from voice_detector import VoiceDetector

detector = VoiceDetector(preset="podcast")
result = detector.analyze("audio.mp3")
print(result.integration_hints.suggested_num_speakers)
```

## Why pip.muid.io?

The full package has large dependencies (PyTorch, audio processing libraries) and is maintained on a private registry for faster iteration. The public PyPI name is reserved to prevent squatting.

## License

MIT
