Sources & licenses
Transcript Studio credits & licenses
Transcript Studio was built for LoschaZ. The speech recognition runs in your browser with open-source software and an open model:
What we built
The workspace, the audio decoding and splitting at quiet moments, the live transcript, the player sync, search, subtitle cue layout and the TXT, SRT, VTT, Word and JSON exports were written for LoschaZ.
Components & licenses
- Transformers.js 4.3.1 by Hugging Face – Apache License 2.0 (license). Served from this website.
- ONNX Runtime Web by Microsoft – MIT License. Its WebAssembly engine is loaded from cdn.jsdelivr.net.
- Whisper speech recognition models by OpenAI – MIT License; the ONNX versions whisper-base and whisper-small are published by the ONNX Community on Hugging Face and downloaded from there.
Our changes
The libraries and models are used unchanged.