Twitxy Saved

pyannoteAI

@pyannoteAI

Make the most of conversational speech with state of the art voice AI
France pyannote.ai Joined April 2024
12 Following 563 Followers 145 Posts

Latest posts 5

pyannoteAI @pyannoteAI
A 100 ms overlap at a turn change. Missed by segmentation, labeled as single-speaker, repeated across thousands of clips.
That's how voice bleed and dropped words end up in a model trained on "clean" data.

➡️ Webinar on Voice AI training data, Nov 24, 5 PM CET. Thread below.
pyannoteAI @pyannoteAI
Hosting audio in an EU region is not the same as keeping it under EU jurisdiction. Our article on sovereign Voice AI covers deployment options, residency versus jurisdiction, and what GDPR, HIPAA, and SOC 2 mean for a voice pipeline.