CLAP2 Audio Segmentation Demo Built on Last Week's Model Release

wightmanr · x · 2026-11-17

Ross Wightman highlighted an audio segmentation demo by Christoph Schuhmann built on the CLAP2 models released last week. The demo was shared on YouTube and is also available as a Hugging Face Space.

CLAP2 is an audio-text contrastive model; the demo shows a practical segmentation use case.

Original post →

More from Multimodal

Multimodal channel →