🎉 Initial Release of FluidAudioSwift
- Speaker Diarization: Automatically identify and separate different speakers in audio recordings
- Speaker Embedding Extraction: Generate speaker fingerprint embeddings for different speakers
- CoreML Integration: Native Apple CoreML backend for optimal performance on Apple Silicon and iOS support
- Real-time Processing: Support for streaming audio processing with minimal latency
- Cross-platform Support: Full support for macOS 13.0+ and iOS 16.0+
Key Features
- On-device speaker diarization with no cloud dependencies
- Configurable clustering thresholds and speaker detection parameters
- Real-time audio processing capabilities
- Comprehensive Swift API with async/await support
Installation
Acknowledgments
This project builds upon the excellent work of the sherpa-onnx project for speaker diarization algorithms and techniques.
*This is the initial release of FluidAudioSwift. We're excited to bring on-device speaker diarization to the Apple ecosystemtag -l