resemble-ai/Resemblyzer
Python package that generates 256-dimensional voice embeddings using deep learning for speaker verification and diarization.
Not currently ranked — collecting fresh signals.
star history
Resemblyzer provides a voice encoder deep learning model that produces compact embedding vectors from audio files, enabling tasks like speaker diarization, speaker verification, and fake speech detection. The embeddings summarize vocal characteristics and can be compared for similarity or clustered for identification. The package includes demos for speaker recognition with limited reference audio and modest fake speech detection.
Frequently asked
- What is resemble-ai/Resemblyzer?
- Python package that generates 256-dimensional voice embeddings using deep learning for speaker verification and diarization.
- Is Resemblyzer open source?
- Yes — resemble-ai/Resemblyzer is open source, released under the Apache-2.0 license.
- What language is Resemblyzer written in?
- resemble-ai/Resemblyzer is primarily written in Python.
- How popular is Resemblyzer?
- resemble-ai/Resemblyzer has 3.3k stars on GitHub.
- Where can I find Resemblyzer?
- resemble-ai/Resemblyzer is on GitHub at https://github.com/resemble-ai/Resemblyzer.