WhisperJAV

Multilingual speech understanding: ASR + emotion recognition + audio event detection. 50+ languages, 15x faster than Whisper, non-autoregressive.

Open-source, accurate and easy-to-use video speech recognition & clipping tool. LLM-based AI clipping integrated.

UQLM: Uncertainty Quantification for Language Models, is a Python package for UQ-based LLM hallucination detection

Fully Open Framework for Democratized Multimodal Training