Can I simplify Inference in this case?

[zuowanbushiwo]

Hi hbredin

Thanks your open source! I read the article pyannote.audio speaker diarization pipeline at VoxSRC 2023. The entire speaker diarization pipeline contains 3 parts: local end-to-end neural speaker segmentation with 10-seconds windows with 1-second stride，neural speaker embedding of each speaker of each window, and agglomerative hierarchical clustering.

If I have a 2-minute audio clip, and I'm sure there are only 3 speakers in the entire clip, Can I remove the speaker embedding and clustering, only use one global end-to-end neural speaker segmentation using the entire speech as input? Will there be any speaker confusion problem if I do this? Or are there other adverse effects?

Thanks！

[github-actions]

Thank you for your issue.
We found the following entries in the FAQ which you may find helpful:

• Can I use gated models (and pipelines) offline?
• Can I apply pretrained pipelines on audio already loaded in memory?

Feel free to close this issue if you found an answer in the FAQ.

If your issue is a feature request, please read this first and update your request accordingly, if needed.

If your issue is a bug report, please provide a minimum reproducible example as a link to a self-contained Google Colab notebook containing everthing needed to reproduce the bug:

• installation
• data preparation
• model download
• etc.

Providing an MRE will increase your chance of getting an answer from the community (either maintainers or other power users).

Companies relying on pyannote.audio in production may contact me via email regarding:

• paid scientific consulting around speaker diarization and speech processing in general;
• custom models and tailored features (via the local tech transfer office).

This is an automated reply, generated by FAQtory

[hbredin]

If I have a 2-minute audio clip, and I'm sure there are only 3 speakers in the entire clip, Can I remove the speaker embedding and clustering, only use one global end-to-end neural speaker segmentation using the entire speech as input? Will there be any speaker confusion problem if I do this? Or are there other adverse effects?

I'd suggest you try and report back :-)

[zuowanbushiwo]

Hi hbredin
The test result of my using this project seems to be OK. This project only uses the segment part, and it is the entire voice input.

[hbredin]

Closing as it reads like the initial question has been answered.