Muse Voice Transcribe
Muse Voice Transcribe is a real time audio perception model that performs streaming speech recognition, speaker diarization and endpointing in a single pass. It ingests audio in 80 millisecond chunks, each turned into one soft token, and handles audio beyond one hour with 20 or more speakers without post processing. It covers 70 or more languages, 25 extensively verified, with native code switching. It reports a 3.1 percent word error rate and a 17.5 percent diarization error rate.
- Type
- Multimodal
- Weights
- Closed
More from Meta Platforms
11
-
Brain2Qwerty v2
Multimodal
End-to-end deep learning pipeline that decodes text in real time from non-invasive magnetoencephalography (ME…
-
Canopy Height Maps v2
Image
Canopy Height Maps v2 (CHMv2) is a global, meter-resolution tree canopy height mapping system built from high…
-
CM3leon
Image
AI model by Meta Platforms
-
Code Llama
Text
AI model by Meta Platforms
-
DexWM (Dexterous Interaction World Model)
Multimodal
DexWM is a dexterous interaction world model for learning hand-object interactions from human videos. It pred…
-
DINOv2
Image
DINOv2 is a self-supervised Vision Transformer developed by Meta AI that learns general-purpose visual featur…
-
DINOv3
Image
DINOv3 is a self supervised computer vision model from Meta that produces universal image backbones without u…
-
Emu Video
Video
AI model by Meta Platforms
About Meta Platforms
1
We're connecting people to what they care about, powering new, meaningful experiences, and advancing the state-of-the-art through open research and accessible tooling.