Whisper ASR

Whisper ASR: OpenAI's robust neural network for accurate, multilingual speech recognition.

Whisper ASR, developed by OpenAI, is an open-source automatic speech recognition (ASR) system. It leverages a large dataset of 680,000 hours of multilingual and multitask supervised data to achieve high accuracy across diverse audio conditions and languages. This deep learning model (a transformer sequence-to-sequence model) excels at transcribing speech, identifying languages, and translating speech into English, making it a versatile tool for developers and researchers.

https://openai.com/research/whisper